THE POTENTIAL OF NOVEL PHENOL-DERIVATIVE COMPOUNDS FOR FEEDING DETERRENT CONTROL OF SEVERAL STORED-PRODUCT COLEOPTERAN PESTS by ERIN CLARK B.Sc., Denison University, 2002 M.Sc., University of Northern British Columbia, 2009 DISSERTATION SUBMITTED IN PARTIAL FUFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN NATURAL RESOURCES AND ENVIRONMENTAL STUDIES UNIVERSITY OF NORTHERN BRITISH COLUMBIA November 2016 © Erin Clark, 2016 Abstract The development of new tactics for the management of stored-product insects is becoming increasingly important. This is due to many factors such as increasing resistance to conventional insecticides making them less effective and the increasing interest in developing chemical control options with fewer negative side effects (e.g., environmental and health impacts). I screened several groups of phenol-derived compounds with varying functional groups for potential as feeding deterrents and insecticides using flour disk, no-choice feeding bioassays against a total of five species of stored-product Coleoptera: Tribolium castaneum, T. confusum, Sitophilus oryzae, S. zeamais, and Rhyzopertha dominica. By using structurally similar compounds, I was able to determine some structure-activity relationships, in particular meta- and para-substituted rings with mid-sized substituents showed the highest feeding deterrent activity. I was also able to show that while there were some similarities in both feeding behavior and mortality between the closely related species, there were also differences. These differences highlight the importance of not extrapolating behavior even to closely related species. There was also evidence that, in general, the primary pests, able to attack whole grain (S. oryzae, S. zeamais, and R. dominica), showed more sensitivity to the test compounds compared to the secondary pests (T. castaneum and T. confusum) which use previously damaged material. This may be explained by looking at these species as relatively specialist and generalist feeders. Finally, during the course of the experiments, an alternative method of measuring feeding in the flour disk bioassays was developed (surface area) and compared to the established method (weight). The methods are comparable and surface area analyses may be a less expensive, alternative way of measuring feeding in some scenarios. Ultimately, I was able to identify ii several compounds that show potential to be feeding deterrents for some stored-product coleopteran pests without significant mortality, which means they may have potential to have a slower rate of resistance development. iii Table of Contents Abstract ........................................................................................................................................... ii List of Figures ................................................................................................................................. v List of Tables ................................................................................................................................. vi Acknowledgement and Dedication .............................................................................................. viii Chapter 1. Introduction ................................................................................................................... 1 Chapter 2. Screening dialkoxybenzenes against Tribolium castaneum (Coleoptera: Tenebrionidae) to detect new feeding deterrents .......................................................................... 17 Chapter 3. Feeding deterrence and toxicity of dialkoxybenzenes to the rice weevil, Sitophilus oryzae (Coleoptera: Curculionidae), and their potential as behavioral control agents. ................ 41 Chapter 4. Feeding choices of five species of stored-product insects to dialkoxybenzenes......... 58 Chapter 5. A comparison of walking behavior of Tribolium castaneum and Sitophilus oryzae in response to seven benzene ring-containing compounds ............................................................. 107 Chapter 6. An inexpensive feeding bioassay technique for stored-product insects .................... 121 Chapter 7. Conclusions ............................................................................................................... 138 Chapter 8. Literature Cited ......................................................................................................... 143 Appendix A. Description of how test compounds were made in Dr. Plettner’s laboratory at Simon Fraser University. ............................................................................................................ 175 Appendix B. Test compounds and their R groups used in a no-choice feeding bioassay on T. castaneum. .................................................................................................................................. 176 Appendix C. Percent feeding by insects on flour disk treated with four doses of various compounds after one and two weeks. ......................................................................................... 179 iv List of Figures Figure 2.1. An example of the percent feeding reduction of T. castaneum on flour disks treated with the test compounds. ........................................................................................................... 29 Figure 3.1. Structure of several compounds that show feeding deterrence bioactivity to Sitophilus adults. ....................................................................................................................... 44 Figure 3.2. Correlation of mean survival and mean amount of disk consumed (±SE). ............ 53 Figure 4.1. Examples of some naturally occuring compounds that show feeding deterrence against some stored-product insects. ......................................................................................... 61 Figure 4.2. Correlation of the percent feeding by S. oryzae after three days and the average percent survival after 14 days (±SE) for all the compounds tested. .......................................... 72 Figure 4.3. Correlation of the percent feeding by S. zeamais after three days and the average percent survival after 14 days (±SE) for all the compounds tested. .......................................... 74 Figure 4.4. Correlation of the percent feeding by T. castaneum after three days and the average percent survival after 14 days (±SE) for all the compounds tested. .......................................... 76 Figure 4.5. Correlation of the percent feeding by T. confusum after three days and the average percent survival after 14 days (±SE) for all the compounds tested. .......................................... 78 Figure 4.6. Correlation of the percent feeding by R. dominica after three days and the average percent survival after 14 days (±SE) for all the compounds tested. .......................................... 80 Figure 5.1. Mean percentage of time spent on either the treated or untreated side of the filter paper based on 15-second observations over five minutes (±SE) by Tribolium castaneum and Sitophilus oryzae. .................................................................................................................... 114 Figure 6.1. A: Photograph of disks ......................................................................................... 126 Figure 6.2. A: Temperature (oC) and percent relative humidity inside the box when first opened on each day. B: Percent of disk remaining (mean ± SE) in each box (room temperature and growth chamber), measured by mass and area, fed on by 25 adult T. castaneum. .......... 129 Figure 6.3. Linear regression (± 95% CI) of the flour disk area (cm2) based on scanning and mass (g) between the disks held at (A) room temperature and (B) held at 30oC. ................... 131 Figure 6.4. Linear regression (± 95% CI) of the flour disk area (cm2) based on photographs and mass (g). ........................................................................................................................... 132 Figure 6.5. Mean percent loss (±SE) of flour disk area (cm2) and mass (g) over four hours at room temperature and humidity (21.1oC, 20% RH)................................................................ 133 v List of Tables Table 2.1. Compounds tested in a no-choice feeding bioassay (bioassay #2) with adult T. castaneum based upon the initial feeding bioassays (bioassay #1). .......................................... 24 Table 2.2. Experimental compounds used in a no-choice feeding bioassay (bioassay #3) with adult T. castaneum..................................................................................................................... 26 Table 2.3. Summary table for the second no-choice feeding bioassay with adult T. castaneum (±SE). ........................................................................................................................................ 31 Table 2.4. Summary table for no-choice feeding bioassay (±SE). ............................................ 32 Table 3.1. Compounds tested in a no-choice feeding bioassay using adult S. oryzae. ............. 46 Table 3.2. Feeding deterrence and toxicity of feeding bioassays by S. oryzae at the high dose of 5 µg/replicate (6 replicates/treatment) after three days. ....................................................... 50 Table 3.3. Feeding deterrence and toxicity results for feeding bioassays by S. oryzae at the low dose of 1.67 µg/replicate (n = 3) after three days. .................................................................... 51 Table 4.1. Abbreviation used to name each experimental compound used in the feeding bioassay. .................................................................................................................................... 65 Table 4.2. Mean weight of Coleopteran species (±SE) tested .................................................. 69 Table 4.3. Feeding and mortality (mean ± SE) of S. oryzae in no-choice feeding bioassay on flour disks with different compounds........................................................................................ 71 Table 4.4. Feeding and mortality (mean ± SE) of S. zeamais in no-choice feeding bioassay on flour disks with different compounds........................................................................................ 73 Table 4.5. Feeding and mortality (mean ± SE) of T. castaneum in no-choice feeding bioassay on flour disks with different compounds................................................................................... 75 Table 4.6. Feeding and mortality (mean ± SE) of T. confusum in no-choice feeding bioassay on flour disks with different compounds........................................................................................ 77 Table 4.8. Summary table for control disk no-choice feeding bioassay (mean ± SE). ............. 83 Table 4.9. Summary table for methanol control disk no-choice feeding bioassay (mean ± SE). ................................................................................................................................................... 84 Table 4.10. Summary table for test compound 3c{2,2} no-choice feeding bioassay (mean ± SE). ............................................................................................................................................ 85 Table 4.11. Summary table for test compound 3c{3,3} no-choice feeding bioassay (mean ± SE). ............................................................................................................................................ 86 Table 4.12. Summary table for test compound 3c{4,4} no-choice feeding bioassay (mean ± SE ................................................................................................................................................... 87 vi Table 4.13. Summary table for test compound 3c{3,6} no-choice feeding bioassay (mean ± SE). ............................................................................................................................................ 88 Table 4.14. Summary table for test compound 3c{n5,6} no-choice feeding bioassay (mean ± SE). ............................................................................................................................................ 89 Table 4.15. Summary table for test compound 3c{n5,n5} no-choice feeding bioassay (mean ± SE). ............................................................................................................................................ 90 Table 4.16. Summary table for test compound 3c{4,n5} no-choice feeding bioassay (mean ± SE). ............................................................................................................................................ 91 Table 4.17. Summary table for test compound DEET no-choice feeding bioassay (mean ± SE). ................................................................................................................................................... 92 Table 4.18. Mean amount consumed (mg/beetle-day) divided by the mean insect weight in the no-choice feeding bioassay at the high treatment dose ............................................................. 93 Table 4.19. Mean amount consumed (mg/beetle-day) divided by the mean insect weight in the no-choice feeding bioassay at the low treatment dose (1/3 the concentration of the high dose). ................................................................................................................................................... 94 Table 5.1. Structures of test compounds used in walking bioassay of Tribolium castaneum and Sitophilus oryzae.. ................................................................................................................... 111 Table 5.2. Mean number of times (±SE) individual insects crossed the dividing line in the middle of the filter paper over the five minutes for both Tribolium castaneum and Sitophilus oryzae. ..................................................................................................................................... 116 Table 5.3. Percentage of insects found on the treated side of the filter paper after fifteen minutes for each species. ......................................................................................................... 117 Table 6.1. The Michaelis-Menten equation Change (%) = [Vmax * Time (h)] / [Km + Time (h)] for the four treatments where Vmax is the asymptote and Km is the Michaelis constant which represents the length of time in which the disk would have lost half the water...................... 134 vii Acknowledgement and Dedication I wish to thank my supervisor, Dr. Dezene Huber, for all of his help and support through this process and my committee members: Dr. Erika Plettner, Dr. Russ Dawson, Dr. Staffan Lindgren, and Dr. Allan Carroll. Without their guidance this thesis would not have been possible. I also wish to thank Dr. Paul Fields and his lab at the Agriculture and Agri-Foods Laboratory in Winnipeg, MB for sharing their knowledge, laboratory, support, and insects. Many people helped with this project and I would like to thank them: Rylee Isitt, Jotvir Mann, Tannis Mayert, Jeanne Robert, Kim Stadnyk, and Yang Yu. None of this research could have been done without financial support from the Natural Sciences and Engineering Research Council of Canada, Canada Research Chairs Program, the Canada Foundation for Innovation, the BC Knowledge Development Fund, and the University of Northern British Columbia. Finally, I wish to thank my family for their support through this entire process, in particular my husband Marcel. It could not have been done without you. viii Chapter 1. Introduction Insects are abundant, exist in almost every ecosystem, and fulfill many ecological roles including decomposition, pollination, predation, and as a food source for other animals. While many insects are beneficial, others are viewed as pests as they threaten resources valued by humans (Foster and Harris, 1997). One area in which insects often cause damage to resources is in stored products. Insects can attack and damage food sources, both raw and processed, at every stage of the marketing system (Hagstrum and Subramanyam, 2006). There are many species of insects that are adapted to use stored products, the majority being members of the order Coleoptera (Hagstrum and Subramanyam, 2006). Estimates vary, but the loss of grain and food products to pests is estimated to be 10-15% annually (Rajendran, 2002). The type of loss associated with insects varies depending on the species but can include contamination (e.g., webbing, fecal matter), destruction of actual commodity including the germ of seeds, and the cost associated with monitoring and management of the insect pests (Hagstrum and Subramanyam, 2006; Mason and McDonough, 2012). In Canada, there is officially a zero tolerance policy for live stored-product insects in grain (Canada Grain Act, 1996). Further, manufacturers must be concerned about stored-product insects entering packaging after it leaves the facility as the consumer may still hold the manufacturer responsible (Hou et al., 2004). This may affect the willingness of the consumer to purchase the product even if the infestation occurred on store shelves or in transit. Stored-product insects are described as being primary or secondary pests. Primary pests are capable of attacking and using intact grains and the larvae often develop inside a protective 1 grain kernel. Examples include Sitophilus spp. (Coleoptera: Curculionidae) and Rhyzopertha dominica (Coleoptera: Bostrichidae). Secondary pests require the grain to be previously damaged or processed to successfully use them. Examples of secondary stored-product insects include Tribolium spp. (Coleoptera: Tenebrionidae) and Cryptolestes spp. (Coleoptera: Laemophloeidae) (Rees, 2004). Control of stored-product insects There are choices when it comes to the control of stored-product insects including physical control (such as by temperature; Fields and Muir, 1995), biological control (Brower et al., 1995), and chemical control. The use of chemicals to control stored-product insects is strictly regulated due to risk of food contamination (White and Leesch, 1995) and the type of chemical used depends on the species being targeted as well as where it is being applied. An insecticide can be defined as a specific form of pesticide formulated to kill an insect at particular life stages. Insecticides can be used as grain protectants (directly applied to the commodity), structural sprays (applied to structures such as the storage bins), surface or spot treatments, aerosols, and fumigants to achieve chemical control (White and Leesch, 1995; Arthur and Subramanyam, 2012). The types of chemical control available for use against stored-product insects have been reviewed extensively elsewhere (White and Leesch, 1996; Hagstrum and Subramanyam, 2006; Arthur and Subramanyam, 2012; Opit et al., 2012; Phillips et al., 2012). Benefits from and drawbacks to the use of insecticides Insecticides are effective against many insect pests, including those that use stored products, and thus are important tools to protect food resources. An increasing world population means an increasing demand for food and fiber, and historically synthetic insecticides have been major contributing factors to steadily increasing the food harvest (Ryan, 2002). Further, 2 insecticides have effectively been used to reduce diseases vectored by insects, such as the reduction of morbidity due to malaria in young children through the use of insecticide-treated nets in Tanzania (Maxwell et al., 2002). There also can arguably be environmental benefits to using pesticides in situations such as the control of invasive species (Myers et al., 2000; Cooper and Dobson, 2007), which can have enormous economic and environmental impacts (Pimentel et al., 2005). So although much attention focuses on the negative aspects of pesticide use, their benefits also must be recognized (reviewed by Cooper and Dobson, 2007). Despite the many benefits of insecticides, there are drawbacks to their extensive use. Insecticides can kill non-target insects (Aebischer, 1990), including beneficial insects such as pollinators. There have also been documented instances where the application of an insecticide to control one pest species has allowed a previously non-important pest to grow to economically damaging levels (Metcalf, 1980; Mochizuki, 2003). This has been attributed to the insecticide eliminating or reducing a predator that was previously keeping the population under control (Metcalf, 1980). Parasite/parasitoid populations also can be significantly decreased by the application of insecticides (and other pesticides) which can allow target insect pest populations to grow unchecked by these natural enemies (Matlock and de la Cruz, 2002; Van den Berg et al., 1998). Further, insects are important food sources for many other animals and insecticide residue can enter the food chain and be consumed by other species such as birds (Morrissey et al., 2007). Additionally, insecticides can leave environmental residues, which can be exacerbated by improper application leading to runoff, etc. Examples include an association between declines in catches of Atlantic salmon, Salmo salar (Salmoniformes: Salmonidae) and control of spruce budworm, Choristoneura fumiferana (Lepidoptera: Tortricidae) with insecticide sprays (Fairchild et al., 1999); terrestrially applied pesticides are often found in aquatic systems (Ritter, 3 1990; Cerejeira et al., 2003). Environmental and health problems in the U.S. attributed to pesticide use were estimated by Pimentel (2005) to have an annual cost of $12 billion dollars. The inequitable distribution of the negative consequences of insecticide use should also be noted. Newer insecticides with advantages such as low mammalian toxicity and lower nontarget effects are termed “reduced risk” insecticides by the United States Environmental Protection Agency (Devine and Furlong, 2007). Their use is growing in developed countries (Devine and Furlong, 2007). However, older or more acutely toxic and environmentally persistent pesticides are still available on the larger world market, despite being banned in ‘western’ countries, and it is those pesticides that agriculturalists in poor nations tend to rely on (Ecobichon, 2001). In addition, there is increasing concern that developing nations lack the regulatory framework or education on the risks and proper application methods of pesticides (Ecobichon, 2001), which can worsen negative consequences of pesticides (Devine and Furlong, 2007). Even within a developed country like the United States, migrant farm workers have more toxicity-related chemical injuries than any other work group, and that demographic is composed primarily of minorities and the socioeconomically poor (Hansen and Donohoe, 2003). There is a disproportionate burden of the costs to the uses of pesticides and insecticides on the poor and less privileged. Insecticide Resistance Another negative consequence of the extensive use of insecticides is development of insecticide resistance in target and non-target insects. The development of insecticide resistance is an increasing problem, with resistance being found in many species of stored-product pests (Fields, 1992; Champ and Dyte, 1977; Subramanyam and Hagstrum, 1995). Resistance is the 4 ability by some individuals to survive a substance at a level that would normally be lethal (Subramanyam and Hagstrum, 1995). The basis of insecticide resistance in insect populations is that the application of the insecticides applies a selective pressure. Some individuals survive despite the presence of an insecticide because they possess genes that confer resistance. Those individuals reproduce following the selective screen of insecticide application, and this enables the genes that provide resistance to be passed on, increasing the proportion of individuals that carry the resistant trait in the population. The number of generations required to develop resistance depends on the selection pressure, the insect species and genetic make-up, and the environment (Subramanyam and Hagstrum, 1995). Most insects have short generation times, however, which means that a population can rapidly display resistance traits in a large proportion of its individuals (Bellinger, 1996). Development of insecticide resistance results in the need for additional applications, higher doses, and the use of different – sometimes more harmful – insecticides, resulting in economic and environmental consequences (Pimentel et al., 1992). For example, the application rate of malathion in Mexico had to be increased several times over the years to protect stored maize and wheat from insects (Perez-Mendoza, 1999). In Brazil, overreliance on historically used insecticides (organophosphates and pyrethroids) led to resistant populations of S. zeamais which directly led to the recommendation and registration of insecticide mixtures for use against them (summarized by Corrêa et al., 2011). In addition, the application of insecticides reduces the population of natural enemies as non-target insects are usually also killed, further releasing the resistant pest population from other types of population controls (Georghiou, 1972). It is estimated that economic losses in the U.S. attributed to insecticide resistance were $400 million in the early 1990s (Pimentel et al., 1992). 5 There are several physiological and behavioral mechanisms that contribute to the evolution of pesticide resistance in insects. One is the ability of the insect to detoxify the insecticides. There are cases in which such insecticide resistance has been traced to a single mutant cytochrome P450 gene (Liu and Scott, 1998; Daborn et al., 2002; Ishak et al., 2016). Cross-resistance can further complicate control methods. Cross-resistance is when the resistance to one insecticide confers resistance to another (even without exposure). For example, sawtoothed grain beetles, Oryzaephilus surinamensis (Coleoptera: Silvanidae), that show resistance to the synthetic insecticide chlorpyrifos-methyl showed increased resistance to essential oil fumigation (Lee, 2002). It was proposed that overall increased levels of various detoxification enzymes explained the resistance to a novel treatment. Significant correlation between resistance to esfenvalerate and to permethrin in S. zeamais supports the concept of cross-resistance (MotaSanchez et al., 2006; Corrêa et al., 2011). Further, some insects have developed resistance to multiple insecticides which, due to past history of insecticide use, enables populations to express resistance to different classes of insecticides with different modes of action (Metcalf, 1980). This prevents the ability of managers to control insects by reverting back to previously used insecticides, further reducing the number of options available (Metcalf, 1980). In addition to metabolic resistance (i.e., modification of the toxin to a less-toxic compound), there is also a possibility for target-site resistance. Many insecticides target aspects of the central nervous system but if pesticide molecules cannot attach to the target site then the pesticide is ineffective. For example, some species of insects are resistant to pyrethroids due to metabolic detoxification (cytochromes P450) (Müller et al., 2008; Wondji et al., 2009) and reduced target-site sensitivity of sodium channels (Dong, 2007). Binding of pyrethroids to 6 sodium channels is the mode of action for this insecticide and mutations in the sodium channels can reduce binding, thus reducing insecticide efficacy (Zhu et al., 2010). Another mechanism of resistance to insecticides includes reducing cuticular penetration of the insecticide. This results when the insect’s cuticle develops barriers that slow the absorption of chemicals. Penetration rates can also vary by life stage, such as malathion more rapidly penetrates the cuticle of immature locusts than mature individuals (Ahmed and Gardiner, 1970). Of course, a combination of various mechanisms can result in increased resistance to insecticides. Insecticide resistance to deltamethrin (a member of the pyrethroid family) in Spodoptera exigua (Lepidoptera: Noctuidae) is due to delayed penetration as well as an increased ability to degrade the insecticide (Delorme et al., 1988). A gene conferring some resistance in the house fly, Musca domestica (Diptera: Muscidae), to DDT and parathion was identified as slowing absorption (Plapp and Hoyer, 1968). Delayed penetration in house flies by itself was not very effective in protecting the insect from pesticides, but in combination with metabolic detoxification, reduced penetration had a large effect on the evolution of insecticide resistance in this species (Sawicki and Lord, 1970). In addition to the detoxification, target site, and reduced penetration as modes of insecticide resistance, there is also the possibility of resistance developing through behavioral traits (Georghiou, 1972; Gould, 1984). Insects can survive by reducing their exposure to insecticide, reducing the likelihood that a lethal dose is ingested or absorbed (Georghiou, 1972). For example, there has been evidence of behavioral avoidance of the organophosphate fenitrothion by some S. zeamais (Braga et al., 2011), and Guedes et al. (2009) found resistant S. zeamais had higher take-off rates when exposed to deltamethrin. This behavior would increase the chance of survival by reducing the length of exposure to the insecticide. 7 There are ways to mitigate the development of insecticide resistance. One technique is to create a refuge where the insects are not exposed to the insecticide (Onstad, 2008). This promotes mating between resistant and susceptible insects, resulting in maintenance of nonresistant alleles in the population. Rotating different treatments in time – i.e., creating a temporal refuge – reduces resistance assuming that there is no cross-resistance to the treatments (Opit et al., 2012). In addition, reducing the applications of insecticides will reduce the strength of selection although this requires the use of other methods of control in addition to insecticides (Opit et al., 2012). The use of mixtures of pesticides with different modes of action can also reduce insecticide resistance (Leeper et al., 1986). However, using an insecticide until it fails and then finding another insecticide is not an effective way to manage insecticide resistance (Hagstrum and Subramanyam, 2006). Behavioral control One potential tactic for insect management other than, or in combination with, the use of insecticides is direct behavioral control. Behavioral control is the manipulation of a pest’s behavior to protect a resource. Trap cropping – stands of plants grown to attract insects to protect the target crop (Hokkanen, 1991; Shelton and Badenes-Perez, 2006) – has been used to protect valued crops and reduce the need for insecticide application (Mitchell et al., 2000). The effectiveness of trap cropping can be further enhanced by the use of behavioral controls such as attractants to draw the pest to the trap crop (Hokkanen, 1991). To use behavioral control as an effective pest management strategy, the behavior of the pest causing the damage (e.g., feeding) is first identified. Then a way to manipulate the behavior must be found, followed by a way to effectively use the manipulated behavior to protect the desired resource (Foster and Harris, 1997). However, in practice, it is the availability of a way to manipulate the behavior that 8 determines which behavior is targeted and how it is targeted (Foster and Harris, 1997). In other words, it is the manager’s ability to change a behavior that directs the use of behavioral control rather than the behavior itself. The targeted behavior may be the action that causes the damage, such as feeding, but it may also be effective to target behavior that is closely related to the damage such as an insect’s searching behavior in relation to the resource (Foster and Harris, 1997). For example, oviposition could be targeted for manipulation as oviposition deterrence could reduce the pest population size or reduce eventual larval feeding. Foster and Harris (1997) identify several attributes that are important for the choice of the stimulant used for behavioral manipulation including, among others: accessibility (the insect must be able to detect the stimulus), specificity (the more specific the stimulus is to the insect species and behavior the more likely it will successfully manipulate the behavior), and practicability (the side effects and/or cost of the stimulus must be within acceptable limits). In some ways, the criteria in selection of a stimulant for behavioral modification, such as being inexpensive and having low non-target toxicity (including to humans), are similar to the criteria for effective insecticides (Isman, 2002; White and Leesch, 2006). It is these attributes of successful stimuli for behavioral manipulation that tend to favor the use of chemicals as stimuli (Foster and Harris, 1997). For a pest manager to effectively manipulate a behavior with a given chemical or other means, a target insect must be able to detect the stimulus. Chemical stimuli (both natural and synthetic) are detected by insects using chemoreceptors. Sensilla, often located on antennae but also other parts of insect bodies, have a pore or pores through which odorant molecules can travel (Steinbrecht, 1997). The molecules are moved through the lymph in the sensillum either by diffusion or by odorant-binding proteins or related pheromone binding proteins to a receptor 9 imbedded in the dendrite membrane inside the sensillum. The odorant binding proteins appear to be the first component of an insect’s olfactory system that can discriminate between odorants. The receptors also discriminate between different odorants, as only suitable receptor types will respond to any given molecule (Jacquin-Joly and Merlin, 2004). The interaction between the molecule and receptor results in a depolarization of the nerve which starts a chain reaction that sends the electrical signal to the insect’s brain, resulting in the appropriate response. Behavioral control may be achieved through the use of a combination of stimuli each evoking different behaviors. An example of this is the push-pull strategy where a repellent stimulus “pushes” the insect away from a resource while at the same time an attractive stimulus “pulls” the insect away from the resource (Pyke et al., 1987; Cook et al., 2007), perhaps even into a lethal trap or toxic situation. Pyke et al. (1987) found that a trap crop “pull” and extract from the Neem tree, Azadiracta indica (Sapindales: Meliaceae), as the “push” stimulus reduced Heliothis (Lepidoptera: Noctuidae) eggs found in the crop to below economic thresholds. While the stimulus may be an attractant or repellent and operate at some distance, the “push” stimulus also could have a short-range effect, such as a gustatory feeding deterrent. Chemical stimuli are important for many insect behaviors. Chemical stimuli modify and direct searching behavior, an essential behavior for location of resources such as food, habitat, and mates (Bell, 1990). Insects detect chemicals to find hosts, avoid non-hosts, locate food sources, communicate, induction of feeding behavior, find mates, and oviposition selection (Wilson, 1965; Paré and Tumlinson, 1999; Sambaraju and Phillips, 2008; Elkinton et al., 1981; Murlis et al., 1992; Bruce et al., 2005). 10 Feeding deterrence One type of behavioral control method that has potential to be an effective tool to control insect pests is feeding deterrence. Frazier and Chyb (1995) discussed three levels where feeding inhibition can occur: preingestion (involves gustatory receptors), ingestion (food transport and activation of salivary enzymes), and postingestion (digestion and absorption of food). The term antifeedant has been used synonymously with feeding deterrent, but a more conservative definition describes antifeedants as those that affect the peripheral sensilla (Isman et al., 1996; Isman, 2002) which, using Frazier and Chyb’s (1995) criteria, would be preingestion. To avoid confusion, I will use the term ‘feeding deterrent’ as I will not be specifying when in the feeding process the deterrence is occurring and deterrence at any stage has potential to be effectively used as a pest management tactic. There are examples of feeding deterrence that are not correlated with toxicity (Cottee et al., 1988; Koul et al., 2004). The choice of food is primarily attributed to chemoreception and, as the mode of action for chemicals that affect feeding behavior is largely unknown, the general avoidance of a food is probably due to chemoreceptors that have broad sensitivity to a spectrum of deterrents (Koul, 2008). There is likely more than one type of receptor involved in feeding deterrence and there are likely different responses to different structures (Simmonds et al., 1990; Koul, 2008). It is also possible that feeding deterrence is not only the result of deterrent receptors but also the stimulation of other receptors that would send signals to activate or reduce feeding. For example, there is evidence that azadirachtin not only stimulates deterrent receptors but appears to suppress sugar receptors (Schoonhoven, 1988 as cited by Koul, 2004), and the CO2 receptors of Drosophila (Diptera: Drosophilidae) can be inhibited by other odorants which have been shown to modify the insect’s behavioral responses (Turner and Ray, 2009). 11 Alternatively, many antifeedants likely have some toxic effect on the insects (Isman, 2002), and the feeding deterrent’s mode of action may be the result of post-ingestive toxicity. Sub-lethal doses of compounds have been shown to reduce feeding (Haynes, 1988; Glendinning, 1996). It has been noted that at an evolutionary level, feeding deterrence is unlikely to continue to result in avoidance unless it is associated with a negative effect on survival (Berenbaum, 1986) and it is logical that from an evolutionary standpoint insects with post-ingestion detection abilities would reduce the likelihood of eating enough to be lethal, increasing their survival rates (Glendinning, 1996). There are limitations and challenges associated with the use of behavioral controls, specifically feeding deterrence. First, significant variation in responses to antifeedants even between closely related species has been observed (Koul, 2008; Akhtar et al., 2008; Simmonds et al., 1990). Second, it has been documented that insects have reduced responsiveness after being exposed to a stimulus (including feeding deterrents). This can happen quite quickly, allowing the insects to feed again in only a few hours (Frazer and Chyb, 1995; Bomford and Isman, 1996). This could be due to habituation (central nervous system level) or due to a sensory adaptation (at the receptor level). Habituation occurs when a repeated stimulus becomes progressively less effective (Mordue et al., 1980) and this has been observed in several insect species either due to adaptation by the receptors or habituation in the central processing system (Cardé and Minks, 1995; Sfara et al., 2011). There is also evidence that habituation to antifeedants is affected by prior exposure (Raffa and Frazier, 1988). This problem of habituation in a pest management context can potentially be mitigated through the use of mixtures (Isman, 2002; Akhtar and Isman, 2003; Koul et al., 2004). In addition, there is evidence that sub-lethal post-ingestive effects can result in a feedback whereby the production of detoxifying enzymes is induced, 12 allowing the insects to continue to feed in spite of the presence of the previously deterrent compound (Lindroth, 1991; Snyder and Glendinning, 1996). Further, one of the problems that may occur with feeding deterrence is that, as discussed with insecticides previously, there is the potential for insects to develop resistance (Koul, 2008). However, resistance to behavioral controls has been less studied than insecticide resistance, probably in part due to the fact that in practice behavioral control is not used as often as insecticides. For an insect to develop resistance, the control method (in this case feeding deterrence) must act as a selective pressure. A strong feeding deterrent may result in the insect staying near the food source and potentially starving, particularly if the insect is unable to move away (non-mobile stage). The insect may have to continue to search for a suitable food source, thus expending energy in continued searching behavior, potentially hurting its chances for survival or reducing its energy stores which could negatively affect its ability to reproduce. Repellents, for example, may keep the pest away from the only suitable resource, causing indirect mortality or lowered reproductive success (Gould, 1984). However, the assumption is that if the control method is non-lethal then the selection pressure is lower which should reduce the rate at which resistance would develop. Finally, it is difficult to connect a stimulus to a behavior. Insects simultaneously receive multiple cues and stimuli, both internally and from their environment, that influence behavior (Riffell et al., 2008; Deither, 1976). Even if only considering feeding, an insect is likely to encounter a mix of compounds in nature (Simmonds et al., 1990). For example, it is thought that host plant selection by insects may be due to the ratio of deterrents and phagostimulants (Chapman, 2003). 13 Summary and objectives It is clear that the development of new methods of stored-product insect control will be important for maintaining food supplies by providing alternative control methods for managers. Such tactics would potentially help to reduce the rates of insect resistance to other control measures, and generally have fewer negative environmental and socioeconomic side effects. My main objective was to test synthetically produced compounds for feeding deterrence against a range of stored-product pests. There is more detail about the stored-product insects being tested in the relevant chapters. Another objective was to identify compounds that may have reduced or no toxicity to the insects, further reducing potential side effects of control methods such as resistance. These compounds can be produced with high purity and have the potential to be made in large quantities (Akhtar et al., 2007). This is in comparison with botanical insecticides which may have more variation in the concentration of active compounds and may be subjected to seasonality (Isman, 2006). There is, however, potential to go further by using compounds for which the structure (i.e., the substituents and their substitution patterns) can be modified in a controlled way and then patterns between the structure and any bioactivity can be detected. Not only could this potentially lead to new control methods but it could expand knowledge about how these chemical structures may be interacting with insect receptors. This could lead to a greater understanding of how these insects are detecting chemical signals in their environment as well as what structures may be most important for maximizing the effects of some chemical control methods. Thus, another objective was to understand the effects of synthetically produced phenol derivatives with varying substitution and substituent patterns in feeding bioassays by several 14 stored-product insects to better understand chemical structural patterns as they relate to behavioral effects. Test compounds Phenols are produced by plants in which they perform many functions including defense against insects and pathogens (Lattanzio et al., 2006). Phenolic odorants and tastants play major roles in the relationship between insects and their environment (Paduraru et al., 2008), and many have been tested for bioactivity against insects, including stored-product pests. For example, eugenol (4-allyl-3-methoxyphenol) and cinnamaldehyde [(2E)-3-phenylprop-2-enal], both phenylpropanoids, are toxic and repellent to Sitophilus spp. (Obeng-Ofori and Reichmuth, 1997; Huang and Ho, 1998; Huang et al., 2002). The phenylpropenes, safrole [5-(2-propenyl)-1,3benzodioxole] and isosafrole [5-[E-prop-1-enyl]-1,3-benzodioxole], were found to deter feeding in adult S. zeamais as well as being toxic to both S. zeamais and T. castaneum (Huang et al., 1999). Dihydroxybenezenes show antifeedant activity against the pine weevil, Hylobius abietis (Coleoptera: Curculionidae) (Borg-Karlson et al., 2006). I therefore predicted that some synthetic phenol derivatives may show similar feeding deterrent, toxic, or repellent effects against the candidate stored-product insects. The common characteristic of phenolic compounds is the presence of at least one hydroxyl–substituted aromatic ring. All of the phenolic test compounds described in this thesis were made as described by Paduraru et al. (2008). Groups of similarly structured test compounds were divided into series identified by letters. For a brief description, please see Appendix A. In the next chapter (Chapter 2) my objective was to screen a large number of synthetic compounds for their potential as feeding deterrants against the stored-product insect, T. castaneum. In addition to trying to assess if any of the compounds reduced feeding, I also 15 determined if the compounds showed any toxicity and if there was a relationship between the structure and the observed bioactivity. As antifeedant effects have a great deal of variability among herbivore species (Koul, 2008; Simmonds et al., 1990; Akhtar et al., 2008), I then tested the synthetic compounds against S. oryzae for feeding deterrence to determine if stronger bioactivity could be detected with a different, but still important, coleopteran stored-product pest (Chapter 3). Again my objectives were to identify any potential feeding deterrent compounds as well as any mortaliy that may or may not be correlated. I also determined if closely related species of stored-product coleopteran pests showed similar resposes to the same compounds. To achieve this, I tested five species of stored-product insects (T. castaneum, T. confusum, S. orzyae, S. zeamais, and R. dominica) against a sub-set of the same potential feeding deterrent compounds (Chapter 4). While stored-product insects’ lifehistories are not often associated with anything other than their close ties to humans and our artifically made stored product environments, I also predicted that a comparison among these species that show different feeding strategies may relate to their responses to my test compounds. The mechanism by which some of the compounds that elicited a feeding deterrent response was unclear (i.e., were the insects not feeding because the compound was repellent or was it due to a gustory or postingestive effect?). To clarify this, I conducted walking bioassays using T. castaneum and S. oryzae (Chapter 5). Finally, during the development of the feeding bioassays, an alternative method of measuring food consumption, using area consumed rather than weight, was investgated to determine if these two methods were comparable (Chapter 6). 16 Chapter 2. Screening dialkoxybenzenes against Tribolium castaneum (Coleoptera: Tenebrionidae) to detect new feeding deterrents Abstract I conducted feeding bioassays using an important stored-product pest, Tribolium castaneum (Coleoptera: Tenebrionidae), in which I screened dialkoxybenzenes for their potential as new feeding deterrents. Flour disks were infused with a number of synthetic phenol-based compounds. The compounds were similar to naturally occurring defensive plant compounds and were tested using multiple no-choice feeding bioassays. In the first bioassays, it was determined that a dose of 100 µg/cm2 was high enough to detect bioactivity and that compounds with metaand para-substutions appeared to be more bioactive. In subsequent feeding bioassays, several of the tested compounds (3c{4,4}, 3c{3,4}, 3c{6,6}) reduced feeding on the flour disks by about 50% or more compared to feeding on the control flour disks after three days. These compounds show promise for development of new methods of commodity protection againts stored-product insect. Introduction The loss of grain and food products to pests, including insects, is estimated to be about 10-15% of the annual crop (Rajendran, 2002). Effective control of stored-product insects can help reduce that loss. Pesticides are extensively used to protect crops and crop products from damaging agents including insects, and it has been estimated that for every dollar spent on pesticide there is a four dollar gain in protected crop (Pimental, 2005). However, along with the 17 benefits to the use of insecticides there are drawbacks. The application of toxic compounds can kill beneficial insects such as pollinators or other insects that are important food for other species or that act as biological control agents (Aebischer, 1990). Insecticides can also enter the wider environment through runoff, ingestion, or other means and result in negative effects on other species including humans. Pimentel (2005) estimates that there is a $10 billion cost annually (in the United States) in environmental and health problems attributable to pesticide use. In addition to these drawbacks, insecticide resistance can further reduce the efficacy of insecticides (reviewed in Chapter 1). These drawbacks have increased the interest in developing alternative strategies to manage and mitigate the damage done by stored-product pests. Behavioral control One alternative strategy for reducing damage from stored-product pests is to use behavioral control. In this method, a behavior of the pest species that is usually related to the damage (e.g., feeding) is identified. Then, the behavior is manipulated in a way to protect the stored product (Foster and Harris, 1997). Certain chemical stimuli may be chosen as behavioral modification agents if they meet certain requirements such as specificity and accessibility as well as proven behavioral activity (Foster and Harris, 1997). Some chemical stimuli can act over long distances and can either repel or attract the insects. Attractants and repellents cause the insect to make oriented movements either towards or away from the stimulant, respectively (Dethier et al., 1960). As an example, sex or aggregation pheromones often are used as a control tactic to attract insects, including some stored-product insects, to lethal or monitoring traps (Lindgren et al., 1985; Thomson et al., 1999; Campbell, 2012). Attractants and repellents can also be used in conjunction as part of a push-pull strategy (Pyke et al., 1987; Cook et al., 2007) where the insects 18 are deterred from one location (where the resource is being protected, the push) with a repellent and simultaneously pulled with an attractant to another location with an attractant stimulus. Alternatively, or additionally, chemical stimuli can act in a short-range manner to induce feeding – often in conjunction with toxins for lethal control – or can serve as deterrents for feeding or oviposition (Foster and Harris, 1997). In this chapter, I tested the efficacy of a number of synthetic compounds as potential feeding deterrents, and their potential lethal effects. In addition, the synthetically produced compounds that I used had varied substiutions which enabled detection of structure-activity relationships. Test Compounds I tested libraries of systematically varied, substituted dialkoxybenzenes, that were phenol derivatives, obtained from the laboratories at Simon Fraser University (Paduraru et al., 2008). Phenols are defensive compounds produced by plants and have been shown to be defensive against insects (Nicholson and Hammerschmidt, 1992; Lattanzio et al., 2006). Substituted phenols have shown antifeedant activity towards Tribolium confusum and Sitophilus granaries (Gabrys et al., 2001). Cinnamaldehyde (a phenylpropanoid), safrole, and isosafrole (phenylpropenes) all show some toxicity to T. castaneum adults but no antifeedant properties (Huang and Ho, 1998; Huang et al., 1999). In extracts from Vernionia oocephala (Asterales: Asteraceae), a shrub native to Africa which contains flavonoids, a group of phenolic compounds – among other compounds – caused feeding inhibition in T. castaneum in laboratory bioassays (Aliyu et al., 2014). I predicted that these test compounds would result in feeding deterrence or mortality in feeding bioassays as they can be included in a class of chemicals known to have 19 some deterrent effects. My goal was to identify compounds that showed feeding deterrence with little or no mortality which could be used as potential tools in a behavioral control strategy. Study species I used adult T. castaneum, the red flour beetle, to assess the bioactivity of the libraries of dialkyoxybenzenes. This insect species is found world-wide and is a major pest of cereals, particularly milled grains. Not a great deal is known about their habitat use before becoming closely associated with humans, but they may have evolved from populations of insects that lived under bark (Lindsley, 1944). Females lay eggs in tunnels in meal products and after the eggs hatch the larvae make their own tunnels as they fed. They then pupate and the adults have been documents living for over 335 days in laboratory conditions (Sinha and Watters, 1985). The insects spend about 66% of their life in the larval stage and 19% as adults (Hagstrum and Subramanyam, 2006). The males produce a pheromone (4,8-dimethyldecanal) (Suzuki, 1981, Faustini et al., 1981) and enantiomers likely serve as an attractant to both sexes (Levinson and Mori, 1983). The pheromone is an effective trap bait (Lindgren et al., 1985; Campbell, 2012). Tribolium castaneum are attracted to grain oils, but the greatest positive chemotaxis is seen with a mixture of pheromone and food volatiles (Phillips et al., 1993). It has also been noted that Tribolium spp. are harder to kill with standard insecticides than are other species of storedproduct insects (Arthur and Subramanyam, 2012), and pesticide-resistant strains of T. castaneum have been identified (Dyte and Blackman, 1970; Zettler, 1991). Further, T. castaneum have had their genome sequenced (Tribolium Genome Sequencing Consortium, 2008). This has allowed for new avenues of research on this insect to further understand its biology as well as its use as a model species. 20 Because it is a stored-product insect found near or within food items, control methods are strictly regulated (White and Leesch, 1996), limiting the types of chemical control tactics that can be used. In addition, stored-product insects not only directly cause loss (i.e., eating the stored product, eating the germ of grain preventing its use as seed), but their presence can result in a product being considered unacceptable by the consumer or by regulatory agencies. Regulations related to insect damage to products (Canada Grain Act, 1996) further increase the economic loss that can result from the insect’s presence. Therefore, developing new control methods - such as behavioral control – would give managers more options and would result in better economic outcomes. Materials and Methods Insects All insects were reared on organic whole wheat flour with 5% (by weight) brewer’s yeast in a growth chamber held at 30oC in darkness. The adult T. castatneum, obtained from laboratory colonies at the Agriculture and Agri-Food Canada labs (Winnipeg, MB), used in all feeding bioassays were under 14 days old at the start of the experiment. Adults, both male and female, were sifted out of the flour and kept in vials in groups of 25 for 24 hours in the dark at 30oC with no food before being placed on the test and control flour disks. Groups of 25 insects were used to ensure that enough flour disks would be eaten to be measureable. Beetles were checked after the 24 hour starvation period to make sure that were still alive, and only insects that appeared visually healthy (e.g., moving around actively, all limbs appeared to intact, etc.) were used in the experiment. 21 Test compounds Test compounds were made following Paduraru et al. (2008; see Appendix A for details). The test compounds used for each bioassay are listed with each bioassay below. In all the bioassays, the individual compounds have a single R1 and a single R2 group. The small libraries are blends of 5 compounds with a single R1 and a range of R2 groups (alkyl = me, et, pr, n-bu, ipent) mixed in equimolar amounts. Feeding bioassay #1 Flour disks were made using unbleached organic white flour, following the methods described in Xie et al. (1996). Test disks were made by adding 200 mg of flour in a 10 mL glass beaker with a magnetic teflon-coated micro stir bar (1/2” x 1/8”, Fisher Scientific) with 950 µL distilled water and 50 μL of methanol:test compound (Appendix B). For the first feeding bioassay, the test compounds (Appendix B) were added in one of four concentrations (1, 50, 100, 200 µg/cm2). The mixture was stirred using the magnetic micro stir bar to create homogenous mixture. Aliquots (100 μL) of the flour mixture were placed on a plastic Petri dish (100 x 15 mm, Fisherbrand, Fisher Scientific) to dry as disks and each disk had a surface area of approximately 0.6 cm2. Once dry, five disks were weighed to an accuracy of 0.001 g using a SI-234 analytical balance (Denver Instruments, Arvada, CO) prior to being placed in a glass Petri dish (100x15mm, VWR International) with 25 adult T. castaneum. A glass Petri dish with 25 beetles but no disks was used as a starvation control. All dishes were placed into a sealed, blacked-out plastic container and kept at 30oC with approximately 50% humidity. At one week and at two weeks, disks were re-weighed as a measure of feeding, and mortality was assessed. For several 22 of the compounds the disks containing the highest concentration of compound stuck to the Petri dish after drying, making it impossible to remove them for use in the bioassay. There were six control no-treatment replicates that were measured at the same time as the treatment plates. Following feeding, treatment disks at the four doses were then compared to the control disks run at the same time. The percent feeding for each treatment at the four doses was assessed and graphed to visualize any trend in amount of feeding and/or a trend in the compound structure and its relationship to bioactivity. A feeding reduction of at least 20% compared to the controls was the threshold used to focus the research on potentially bioactive compounds for subsequent feeding bioassays. Feeding bioassay #2 For the second feeding bioassay, disks were made and insects were tested with controls as described above. However, because I had found that the disks, particularly with higher concentrations of test compound, had a tendency to stick to the plastic petri dish after they had dried overnight, making it difficult to remove them for the feeding bioassay, I used aluminum weigh boats to create disks, which prevented any further difficulties. Rather than test multiple doses, all treatments had a dose of 100 µg/cm2 per replicate (n=3), a dose based off the first feeding bioassay (Table 2.1). I also chose to weigh the disks after three days rather than waiting a week as this was more consistent with feeding bioassays conducted at other laboratories. The no-choice bioassay was set up as described above and the weight of the disks was taken in the same way. Data were analyzed using analysis of variance (ANOVA) and, if there were significant differences, was followed by a Tukey HSD post-hoc analysis if the data met requirements of normality (Shapiro-Wilk) and equal variance; otherwise a Kruskal-Wallis 23 Table 2.1. Compounds tested in a no-choice feeding bioassay (bioassay #2) with adult T. castaneum based upon the initial feeding bioassays (bioassay #1). Compound R1 R2 3c {4,4} n-butyl n-butyl 3c {6,6} allyl allyl 3c {3,n5} propyl n-pentyl 3c {n5,6} n-pentyl allyl 3c {4,6} n-butyl allyl 3c {4,n5} n-butyl n-pentyl 3c {n5,n5} n-pentyl n-pentyl 24 ANOVA on ranks followed by post-hoc tests using Dunn’s method was used. Data for the length of time that it took 50% of the insects to die (LT50) were analyzed using a Kaplin Meier Log Rank survival analysis followed by a pairwise multiple comparison procedure (Holm-Sidak method). All were analyzed using SigmaPlot 12.5. Feeding bioassay #3 Based on the previous feeding bioassays, another no-choice feeding assay was run using compounds that were similar in structure to those that showed some feeding deterrent potential as well as several new ones that were synthesized by our partner lab (Dr. Erika Plettner, Simon Fraser University) after my initial experiments (Table 2.2). I also retested compound 3c{6,6} as it showed an unexplainable negative consumption. I also retested several of the compounds that previously had stuck to the Petri dishes and were not able to be used in feeding bioassay #1 (Appendix C provides results from bioassay #1). Neem, Azadirachta indica, oil was also run as a positive control as it has both antifeedant and insecticidal properties to a variety of insects (Morgan, 2009). Again, disks were made as described above and treatments were all at a dose of 100 µg/cm2 per replicate (n = 6). The experiment was set-up as previously described and disk weight was taken after 3, 7 and 14 days, with mortality checked every day. I checked the disk weight later in the experiment to determine if any effects on feeding could be detected after longer exposure to the test compounds. Data were analyzed as described above. 25 Table 2.2. Experimental compounds used in a no-choice feeding bioassay (bioassay #3) with adult T. castaneum. Individual compounds have a single R1 and a single R2 group; small libraries are blends of 5 compounds with a single R1 and a range of R2 groups (alkyl = me, et, pr, n-bu, ipent). Treatment R1 R2 Compound name Control -- -- -- 3b{1,1-5} alkyl me 1-alkoxy-3-methoxybenzene 3b{2,1-5} alkyl et 1-alkoxy-3-ethoxybenzene 3b{3,1-5} alkyl pr 1-alkoxy-3-propoxybenzene 3b{4,1-5} alkyl n-bu 1-alkoxy-3-butoxybenzene 3b{5,1-5} alkyl i-pent 1-alkoxy-3-isopentyloxybenzene 3b{6,1-5} alkyl allyl 1-alkoxy-3-allyloxybenzene 3b{2,2} et et 1,3-diethoxybenzene 3b{3,3} pr pr 1,3-dipropoxybenzene 3b{4,4} n-bu n-bu 1,3-dibutoxybenzene 3b{5,5} i-pent i-pent 1,3-di-isopentoxybenzene 3b{6,6} allyl allyl 1,3-diallyloxybenzene 3b{1,6} me allyl 1-allyloxy-3-methoxybenzene 3b{2,6} et allyl 1-ethoxy-3-methoxybenzene 3b{3,6} pr allyl 1-allyloxy-3-propoxybenzene 3b{4,6} n-bu allyl 1-allyloxy-3-butoxybenzene 3b{5,6} i-pent allyl 1-allyloxy-3-isopentyloxybenzene 26 3c{2,2} et et 1,4-diethoxybenzene 3c{3,3} pr pr 1,4-dipropoxybenzene 3c{4,4} n-bu n-bu 1,4-dibutoxybenzene 3c{3,4} pr n-bu 1-butoxy-4-propoxybenzene 3c{5,5} i-pent i-pent 1,4-dipentoxybenzene 3c{6,6} allyl allyl 1,4-diallyloxybenzene 3c{3,6} pr allyl 1-allyloxy-4-propoxybenzene 5a{2,1-5} alkyl et 1-allyl-2-ethoxy-3-alkoxybenzene 5a{4,1-5} alkyl n-bu 1-allyl-2-butoxy-3-alkoxybenzene 5a{6,1-5} alkyl allyl 1-allyl-2-allyloxy-3-alkoxybenzene Butyl eugenol -- -- 1-allyl-3-methoxy-4butoxybenzene Neem -- -- Active compound: azadirachtin A 27 Results Feeding bioassay #1 A few of the test compounds showed some feeding deterrence or feeding stimulation compared to the control, although the majority of compounds did not reduce feeding by over 20% compared to controls (Appendix C, Figure 2.1). 3c{4,4} reduced feeding by over 50% compared to controls at the two higher doses tested (100 and 200 µg/cm2). There was very little mortality in any of the treatments at any of the doses tested (Appendix B), except on the disks containing 3b{2,2}, 3b{3,5}, 3b{1,6}, although fewer than half the beetles died. However, there was some variation in the feeding on the disks treated with those (3b{2,2}, 3b{3,5}, 3b{1,6}) compounds and so retesting was necessary. Based on the trends observed, there was more feeding bioactivity on para- and meta-substituted compounds (3b and 3c), and the only mortality observed was with insects feeding on 3b compounds. Therefore, for the next bioassay (#2) the compounds tested were limited to those with para- and meta-substitutions. Based on these results I also determined that a dose of 100 µg/cm2 was sufficient to detect bioactivity while conserving the amount of each test compound being used. Feeding bioassay #2 The results of the second round of no-choice feeding bioassay showed significant variation in the inhibition of feeding by the compounds (difference in disk weight: F7,16 = 18.0, P < 0.001; percent feeding: F7,16 = 19.1, P < 0.001) and post-hoc tests revealed two compounds that significantly reduced feeding during the first three days (3c{6,6} and 3c{4,n5}) as well as several compounds that inhibited feeding to a slightly lesser extent (3c{4,6}, 3c{n5,n5}, and 3c{n5,6}). 28 Figure 2.1. An example of the percent consumption relative to the controls of T. castaneum on flour disks treated with the test compounds. Percent consumed of disks treated with compound 3c{4,4} at four doses (1, 50, 100, 200 µg/cm2) at one and two weeks. Only one replicate for each dose was run. 29 This effect was not as strong after 14 days of feeding (difference in disk weight: F7,16 = 4.20, P = 0.01; percent feeding: H7 = 14.97, P = 0.04) (Table 2.3). There was an increase in weight for the 3c{6,6} disks which did not make sense and should be treated with caution as it could be due to measurement error or an issue with humidity in a subset of the experiment. There were no significant differences in insect survival among any of the treatments compared to the controls (P > 0.05). However, the LT50 that was seen in the treatments 3c{6,6} and 3c{4,6} was around ten days, so the lethal effect was not immediate (Table 2.3). Feeding bioassay #3: There were significant differences between treatments and post-hoc tests showed that feeding on some of the compounds was significantly different from each other but none were different from the controls (Table 2.4). On day 3 and day 7 there were significant differences in insect feeding both when analyzing the data using either the difference in disk weight (difference day 3: H28 = 109.7, P < 0.001; difference day 7: F28,145 = 4.66, P < 0.001) or difference in feeding as a percent of the feeding on the control disks (percent feeding reduction day 3: H28 = 110.55, P < 0.001; percent feeding reduction day 7: F28, 145 = 4.70, P < 0.001). There were significant differences in feeding after 14 days (difference day 14 = F28, 145 = 2.87, P < 0.001; percent feeding reduction day 7: F28, 145 = 2.92, P < 0.001). The compound that showed the highest level of feeding deterrence was 3c{4,4} by both measures and over the two weeks. Several of the same compounds that had been tested in feeding bioassay #3 did not show feeding deterrence as strongly as was detected in the previous feeding bioassay (e.g., 3c{6,6}, 3c{4,6}). There were no significant differences in insect survival after 14 days in any of the treatments (H28 = 30.84, P = 0.32). 30 Table 2.3. Results of no-choice feeding bioassay with adult T. castaneum (means and % ± SE). Treatment doses were 100 µg/cm2 per replicate (n=3). Percent feeding for the control plates was set at 100%. The time for 50% of the insects to die (LT50) could not be calculated for some of the treatments as there was not enough mortality. Significant differences in columns (P < 0.05) are represented by different letters. Treatment Mean amount consumed after 3 days (mg) % Feeding after Mean amount % Feeding after Mean amount 3 days consumed after 3 days (per consumed after 3 days beetle day) 14 days (mg) (mg/beetle day) % Feeding after Survival (# 14 days alive 14 days/beetles day 0) LT50 Control 22.1 (1.5)a 100.0a 0.30 (0.02)a 100.0a 51.83 (1.04)a 100.0 1.00 (0.00) --a MeOH 15.3 (2.0)ab 69.4 (9.2)ab 0.20 (0.03)ab 69.4 (9.2)ab 50.13 (0.72)ab 96.7 (1.4) 1.00 (0.00) --a 3c{6,6} -3.9 (2.8)d -17.6 (12.5)d -0.06 (0.04)d -19.6 (13.7)d 37.93 (6.30)b 73.2 (12.2) 0.64 (0.09) 10.2 (0.6)b 3c{3,n5} 8.8 (1.6)bc 39.8 (7.3)bc 0.12 (0.02)bc 12.7 (7.3)bc 50.83 (2.68)ab 98.1 (5.2) 0.99 (0.01) 13.9 (0.0)a 3c{n5,6} 4.1 (3.3)cd 18.7 (14.8)cd 0.06 (0.04)cd 18.8 (14.8)cd 46.53 (1.78)ab 89.8 (3.4) 0.97 (0.03) 13.7 (0.3)a 3c{4,6} 0.7 (2.4)cd 3.3 (6.2)cd 0.01 (0.02)cd 3.8 (6.9)cd 37.73 (1.47)b 72.8 (2.8) 0.74 (0.05) 11.1 (0.6)b 3c{4,n5} -1.8 (2.1)d -8.0 (9.4)d -0.02 (0.03)cd -8.3 (9.8)d 48.17 (2.09)ab 92.9 (4.0) 0.97 (0.01) 13.7 (0.3)a 3c{n5,n5} 0.3 (1.8)cd 1.3 (8.0)cd 0.00 (0.02)cd 1.2 (8.0)cd 50.30 (1.92)ab 97.0 (3.7) 1.00 (0.00) --a 31 Table 2.4. Results of no-choice feeding bioassay (means and % ± SE). Treatments were all used at a rate of 100 µg/cm2 per replicate (n=6). Percent feeding for the control plates was set at 100%. Significant differences in columns (P < 0.05) are represented by different letters. Treatment Mean amount consumed after 3 days (mg) % Feeding after Mean amount 3 days consumed after 7 days (mg) % Feeding after Mean amount 7 days consumed after 14 days (mg) % Feeding after Survival (# 14 days alive 14 days/day 0) Control 23.9 (1.5)abcd 100.0abcd 44.2 (1.2)abcd 100.0abcd 65.1 (2.2)abc 100.0abc 0.99 (0.01) 3b{1,1-5} 25.3 (0.7)abc 106.2 (2.7)ab 43.4 (0.7) abcde 98.3 (1.5)abcde 65.8 (0.6)abc 101.0 (0.9)abc 0.99 (0.01) 3b{2,1-5} 27.9 (1.7)a 116.8 (7.1)a 48.4 (2.5)a 108.9 (5.7)a 72.3 (3.2)a 111.1 (5.0)a 0.99 (0.01) 3b{3,1-5} 26.4 (1.4)ab 110.7 (5.7)abc 46.8 (1.5)ab 105.8 (3.4)ab 69.7 (2.1)ab 107.0 (3.2)ab 0.98 (0.01) 3b{4,1-5} 26.8 (2.6)abc 112.2 (10.9)abc 45.4 (2.5)abc 102.7 (5.6)ab 70.0 (3.3)ab 107.4 (5.1)ab 0.99 (0.01) 3b{5,1-5} 21.3 (2.1)abcd 89.4 (8.8)abcd 45.0 (3.5)abcd 101.9 (7.9)abc 65.0 (4.3)abc 99.8 (6.6)abc 0.98 (0.02) 3b{6,1-5} 24.3 (1.8)abcd 101.7 (7.7)abcd 45.0 (3.2)abcd 101.9 (7.2)abc 69.3 (3.5)ab 106.4 (5.4)ab 0.99 (0.01) 3b{2,2} 22.9 (2.8)abcd 96.1 (11.6)abcd 40.7 (3.2) abcde 92.2 (7.1)abcde 65.7 (4.9)abc 100.8 (7.5)abc 1.00 (0.00) 3b{3,3} 24.0 (1.6)abcd 100.6 (6.9)abcd 43.8 (2.9)abcde 99.2 (6.6)abcde 66.8 (3.7)ab 102.6 (5.7)ab 1.00 (0.00) 3b{4,4} 22.8 (1.5)abcd 95.7 (6.5)abcd 41.1 (2.5) abcde 93.1 (5.7)abcde 65.1 (3.2)abc 99.9 (4.9)abc 0.99 (0.01) 3b{5,5} 26.7 (0.8)a 111.9 (3.5)a 43.5 (1.1) abcde 98.5 (2.5)abcde 65.5 (1.9)abc 100.5 (2.8)abc 0.99 (0.01) 3b{6,6} 20.4 (1.1)abcd 85.4 (4.6)abcd 39.8 (1.9) abcde 90.1 (4.2)abcde 61.1 (2.4)abc 93.7 (3.6)abc 1.00 (0.00) 3b{1,6} 23.0 (1.3)abcd 96.3 (5.4)abcd 41.5 (2.3) abcde 93.9 (5.2)abcde 65.9 (1.7)abc 101.2 (2.7)ab 0.99 (0.01) 32 3b{2,6} 23.6 (1.6)abcd 98.8 (6.6)abcd 41.1 (1.9) abcde 93.1 (4.2)abcde 63.3 (2.0)abc 97.2 (3.1)abc 0.99 (0.01) 3b{3,6} 22.6 (0.4)abcd 94.9 (1.5)abcd 35.5 (0.8)bcdef 80.3 (1.8)bcdef 56.0 (0.9)abc 85.9 (1.3)abc 0.99 (0.01) 3b{4,6} 25.7 (0.7)ab 107.9 (3.0)ab 36.4 (1.7) abcdef 82.3 (3.8)abcdef 55.8 (3.9)abc 85.6 (6.0)abc 0.99 (0.01) 3b{5,6} 16.3 (1.5)abcd 68.5 (6.2)abcd 35.4 (1.5)cdef 80.2 (3.3)bcdef 59.0 (1.8)abc 90.5 (2.7)abc 0.99 (0.01) 3c{2,2} 15.8 (1.7)abcd 66.0 (7.0)abcd 37.6 (2.4) abcdef 85.2 (5.3)abcdef 61.6 (2.9)abc 94.6 (4.5)abc 0.99 (0.01) 3c{3,3} 14.1 (0.9)bcd 59.1 (3.7)bcd 39.0 (2.6) abcde 88.2 (5.8)abcde 66.1 (4.7)ab 101.5 (7.2)ab 0.98 (0.01) 3c{4,4} 8.1 (1.6)d 33.8 (6.6)d 26.8 (1.2)f 60.6 (2.8)f 49.0 (2.3)c 75.2 (3.5)c 0.99 (0.01) 3c{3,4} 12.9 (1.3)bcd 54.2 (5.3)bcd 32.3 (2.1)ef 73.1 (4.8)ef 58.6 (4.6)abc 89.9 (7.1)abc 0.83 (0.10) 3c{5,5} 21.9 (2.0)abcd 91.7 (8.2)abcd 42.1 (2.2) abcde 95.2 (4.9)abcde 63.6 (4.4) abc 97.6 (6.8)abc 0.98 (0.01) 3c{6,6} 10.4 (2.4)cd 43.5 (10.0)cd 33.2 (3.0)def 75.2 (6.9)def 53.5 (4.8)bc 82.1 (7.3)bc 0.86 (0.08) 3c{3,6} 15.1 (0.7)abcd 63.5 (3.1)abcd 37.7 (1.7) abcdef 85.3 (3.9)abcdef 62.4 (1.9) abc 95.8 (3.0)abc 0.99 (0.01) 5a{2,1-5} 23.5 (1.3)abcd 98.5 (5.4)abcd 33.6 (1.9)cdef 76.0 (4.2)cdef 57.3 (2.5) abc 88.0 (3.8)abc 0.98 (0.01) 5a{4,1-5} 27.9 (2.0)a 116.9 (8.4)a 41.2 (3.5) abcde 93.2 (7.8)abcde 65.2 (4.3) abc 100.0 (6.6)abc 0.99 (0.01) 5a{6,1-5} 24.7 (1.2)abcd 103.4 (4.8)abcd 40.2 (1.7) abcde 91.0 (3.9)abcde 61.8 (2.1) abc 94.9 (3.2)abc 0.98 (0.01) Butyl eugenol 22.6 (1.0)abcd 94.6 (4.4)abcd 39.5 (2.3) abcde 89.3 (5.2)abcde 62.8 (2.7) abc 96.4 (4.1)abc 0.99 (0.01) Neem 16.2 (2.0)abcd 68.1 (8.5)abcd 37.0 (2.2) abcdef 83.6 (4.9)abcdef 53.7 (4.2)bc 82.5 (6.4)bc 0.97 (0.02) 33 Discussion The compounds that elicited the largest feeding reductions after three days were all parasubstituted: 3c{4,4} (Table 2.4), 3c{6,6}, 3c{3,n5}, 3c{n5,6}, 3c{4,6}, 3c{4,n5}, and 3c{n5,n5} (Table 2.2). None of the para-substituted compounds in this study elicited antennal responses in gypsy moths (Lymantria dispar) in another study (Paduraru et al., 2008), but did show inhibition when used in combination with the sex pheromone. Several of the para-substituted compounds elicited strong feeding and oviposition deterrence in Trichoplusia ni (cabbage looper), both with toxicity (e.g., 3c{2,2}, 3c{2,1-5}, 3c{4,1-5}), and without [e.g., 3c{4,4}, 3c{6,6}, 3c{4,6}, 3c{6,1-5}] (Akhtar et al., 2007). I did observe some overlap in effect between the moths and T. castaneum, particularly 3c{4,4}, 3c{6,6}, and 3c{4,6} where feeding deterrence with low mortality was observed in both Trichoplusia ni (Akhtar et al., 2007) and Tribolium castaneum (Table 2.4). Despite some overlap between T. ni and T. castaneum there were many more compounds that showed bioactivity in the former than I observed in the latter (Akhtar et al., 2007; Akhtar et al., 2010). This emphasizes that there are substantially different behavioral responses among insect orders to the same compounds. Antifeedants seem to have more variation in activity than is observed with insecticides (Isman, 2002; Isman, 2006) and even closely related species show significant variation in responses to feeding deterrence (Chapter 3 of this thesis, Isman, 1993; Akhtar et al., 2008). This is important to consider for control strategies as it is unlikely that a single compound will effectively modify all insect pests’ behavior in a resource or location and successfully protect a resource under attack by a number of species. Some compounds that might 34 be an antifeedant to one species could serve as a stimulant for another (Isman, 2006), reducing its effectiveness as a control tactic in some contexts. Strategies to control stored-product insects face the challenge of trying to manage not just one species but the multiple species that coexist in a given stored-product context (Athanassiou et al., 2014). There can be many species of insects from different orders, with different ecological niches, all living in the same storage environment (Arbogast and Throne, 1997; Athanassiou et al., 2005). For example, in stored wheat, T. castaneum, Cryptolestes ferrugineus (rusty grain beetle), and Rhyzopertha dominica (lesser grain borer) were the most common insects found (Athanassiou et al., 2005). All three of these species can do significant damage to grain, so effective control strategies need to address more than just one of them in some instances. My results showed different responses compared to previously tested species (Akhtar et al., 2007; Paduraru et al., 2008; Akhtar et al., 2010). Despite this challenge some of my test compounds may still prove useful as potential tools for tactics in an integrated pest management strategy which combines many techniques, including the potential use of insecticides, depending on the stored-product pests present. There are challenges to the use of antifeedants at an operational level. First, there are requirements for regulations regarding the use of any synthetic compounds around food products. Second, there is the interspecific variation in response to feeding deterrents, discussed above. Third, there is evidence that insects show behavioral plasticity – insects can rapidly habituate to feeding deterrents. This results in the deterrent becoming ineffective, sometimes in a matter of hours (Isman, 2006, Bomford and Isman, 1996, Akhtar et al., 2003). An insect that can easily leave in search of other food may do so upon first encounter but an insect that cannot 35 would remain until the deterrent is no longer effective (Isman, 2006). Many life stages of the stored-product insects are not capable of migration, such as the larval stage of T. castaneum, but still do extensive damage to the product. These insects would remain continually exposed to the antifeedant stimulus until the stimulus may no longer be effective, allowing the insect to continue to feed. In my study, there is a possibility that habituation to test compounds occurred as the insects were exposed for two weeks. Any significant feeding deterrence that was observed in this study was found during the first week and the effect was not observable by 14 days. Habituation is one possible explanation for this observation. It is also possible that the compounds resulted in a sub-lethal toxic effect, discouraging feeding after a small amount was initially eaten. This would mean these compounds were not antifeedants as defined by Isman et al. (2006). These authors defined antifeedants as compounds that deter feeding as a result of direct action on the taste sensilla; however sub-lethal toxic effects can still be an effective tool for pest management (Foster and Harris, 1997). As there was almost no mortality observed in any of the feeding bioassays these compounds clearly do not kill at the doses tested. However, a sub-lethal toxic effect could result in feeding deterrence but not due to a stimulation of deterrent receptors. Sub-lethal doses of insecticides have been shown to decrease feeding in some insects (Haynes, 1988). After the initial feeding deterrence due to this sub-lethal toxic effect, T. castaneum could have increased levels of detoxifying enzymes to enable them to continue feeding on the compound-treated flour disks. Alternately, as it is known that some of the test compounds are volatile (Ebrahimi et al., 2013), the dosage level in the flour disk could have been reduced over time to levels that no longer had an effect on the insects. 36 A further challenge to the use of antifeedants at an operational level is that, like insecticide resistance discussed above, there is a possibility of insects developing resistance to these kinds of stimuli, particularly if used indiscriminately (Koul, 2008). Myzus persicae (Hemiptera: Aphididae), the green peach aphid, showed a nine-fold increase in resistance to azadirachtin (LC50) after 40 generations of being treated with the compound (Feng and Isman, 1995). Mosquitos were selected for “irritability” behavioral responses to DDT in the laboratory, and researchers were able to breed strains that were significantly either more or less irritated by DDT than the original insect population (Georghiou, 1972). Any compound that results in the insect having to expend more energy searching for suitable food sources has the potential to result in selection for resistance. Despite this, the lack of toxicity may result in a lower selection pressure, thus reducing the rate at which resistance to this type of control may develop. Blends can help reduce the development of resistance (Feng and Isman, 1995) as well as reduce the challenge of insects becoming habituated to the antifeedant as discussed above (Isman, 2002; Koul, 2008; Koul et al., 2004). This is one reason I tested base compounds which contained one constant R group and one R group with variations (e.g., 3a{1,1-5}) but none of the tested compounds showed any reduction in insect feeding, even 3c{4,1-5} which retains the most similar structural aspects of the most effective feeding deterrent 3c{4,4}. Test compound 3c{3,4} showed feeding reduction at day 3 which was not significantly different than the feeding reduction by 3c{4,4} (54.2 and 33.8% feeding reduction, respectivly). When three structurally similar compounds taken from A. indica were tested, they all showed antifeedant properties but no potentiating effect, indicating that these structurally similar compounds were competing for the same target site (Koul et al., 2004). 37 Some of the structurally similar compounds that I tested (3c{3,4} and 3c{4,4}) both caused feeding deterrence, which may be an indication that they have similar modes of action and are targeting the same sensory binding sites. However, lack of bioactivity in the mixture (3c{4,1-5}) could indicate that there is an antagonistic relationship in terms of bioactivty between some of the individual compounds in the mixture. This could be caused by individual compounds binding at two different sites which cause opposite effects or by a change in the resulting bioactivity if two compounds bind together. Alternatively, if the other individual compounds in the mixture do not show the same bioactivity, perhaps the mixture had a very low amount of the bioactive compound, below the necessary threshold to result in a detectable feeding deterrence. Future work Compounds that inhibit attraction to aggregation pheromone could also be useful as a management strategy. Trap catches of the European spruce bark beetle, Ips typographus (Coleoptera: Curculionidae), baited with pheromone were significantly reduced by also baiting the traps with antennally-active non-host volatiles (Zhang and Schylter, 2003). Individual ponderosa pine, Pinus ponderosa (Pinales: Pinaceae), have successfully been protected from attacks by western pine beetle, Dendroctonus brevicomis (Coleoptera : Curculionidae) and red turpentine beetle (D. valens) using verbenone and non-host volatiles (Fettig et al., 2008). If some of the compounds I tested reduce attraction of T. castaneum to their sex-pheromone, they could be an effective way of reducing the population through disrupting mating, and thus the damage, caused by large populations. 38 All the bioassays in my study were no-choice, which meant the insects needed to eat the food treated with the compound or starve. This would have strongly motivated the insects to feed, despite possible underlying bioactivity of the compounds. Further work using a choice bioassay could detect a preference by the insect not to eat the compounds if there is an alternative food source. In addition, the compounds could have a repellent effect, which would result in the insect moving away from the compound, something that the insects could not do easily in the sealed Petri dishes. Adult T. castaneum are quite mobile and will disperse from patches of flour (Naylor, 1961; Hagstrum and Gilbert, 1976). This could also be an effective management strategy as discussed previously. Testing for repellency, perhaps using a pitfall bioassay where the insects are presented with two stimuli (a control and a test compound), and where the number of insects that respond to each treatment is used as the measure of attraction or repellency (Phillips et al., 1993; Germinara et al., 2007), could help to elucidate any repellent or attractive bioactivity from these compounds. Conclusion Using a screening feeding assay I was able to narrow down the structure of the test compounds that seem to elicit some feeding reduction (para-substitution, intermediate group sizes), with one compound (3c{4,4}) reducing feeding by almost 70% compared to the controls. None of the compounds tested resulted in detectable levels of mortality with this species. This does support the prediction that a subset of these phenolic-based structures have potential to be used as feeding deterrents, as suggested by similarly structured naturally occuring compounds. The lack of mortality also indicates that there is the potential that these compounds are non-lethal to T. castaneum, thus possibly lowering the rate of developing resistance. However, compared to 39 bioactivity elicited in T. ni (Akhtar et al., 2007; Akhtar et al., 2010), these compounds show a very low level of feeding reduction activity for T. castaneum. Therefore, these compounds should be tested with other species of stored-product insects to determine if the compounds show potential to reduce feeding by other species and to be able to compare T. castaneum responses with those of other beetles (see Chapters 3 and 4). 40 Chapter 3. Feeding deterrence and toxicity of dialkoxybenzenes to the rice weevil, Sitophilus oryzae (Coleoptera: Curculionidae), and their potential as behavioral control agents Abstract The rice weevil, Sitophilus oryzae (Coleoptera: Curculionidae), was used to test the feeding deterrence and toxicity of a group of synthetic phenol derivatives. Flour disks were made and treated with one of two doses of ten dialkoxybenzenes, upon which S. oryzae were allowed to feed in a no-choice bioassay. Feeding on the flour disks was measured at three days and mortality was observed up to 14 days. Several of the compounds tested showed feeding reduction activity, in some cases significantly better than DEET which was used as the positive control. Compounds substituted with longer chains resulted in higher levels of feeding deterrence but were also more toxic to the insects. In general, there was a strong correlation between feeding deterrence for S. oryzae and observed mortality. Introduction Sitophilus oryzae (Coleoptera: Curculionidae), the rice weevil, is a pest of stored cereal and is found worldwide, particularly in warmer regions (Rees, 1996). It, along with S. zeamais and S. granaruis, are probably the most destructive primary pests of stored cereals (Rees, 1996). Sitophilus oryzae are often found in small-grained cereal such as rice or wheat and they can attack processed products such as pasta (Haines, 1981; Rees, 1996). Eggs are laid in a grain and the larvae develop inside it, using the grain as their food source. After completing metamorphosis, an adult S. oryzae will chew its way out of the grain and is then capable of flight dispersal to find a new food source where it can mate and lay eggs (Rees, 1996). Because 41 weevils develop inside the grain kernel, they can be very difficult to kill (White and Leesch, 1995), as they are not accessible by conventional chemical-based control measures. Sitophilus oryzae of both sexes respond to a male-produced aggregation pheromone, sitophilure [(R*,S*)-4methyl-5-hydroxy-3-heptanone] (Phillips et al., 1985; Levinson et al., 1990). Pheromone lures can be used effectivly by managers for detection, monitoring, and trapping of stored-product insects, including S. oryzae (Burkholder and Ma, 1985; Trematerra and Girgenti, 1989; Likhayo and Hodges, 2000). Protection of wheat from S. oryzae (and other stored-product pests) is usually done with a combination of monitoring, sanitation, grain drying, and a variety of pesticides (Hagstrum et al., 1999). However, there is interest in the development of chemical methods that modify behavior to reduce or eliminate feeding. Successful behavioral control should reduce feeding damage to below economic thresholds, but would not necessarily increase mortality of the insects. Advantages to this type of control compared with traditional pesticides is a reduction in toxic compounds entering the food chain and populations developing resistance. Pesticides have been linked to cancers in humans and there are studies on animals suggesting that pesticides can result in immune dysfunction (reviewed by Pimentel et al., 1992). The increased use of insecticides in rice production in Indonesia caused the destruction of the natural enemies of one type of rice pest, resulting in a population explosion and loss of rice (Pimentel et al., 1992). There are field populations of Sitophilus spp. that have been shown to have resistance to a range of insecticides including lindane, malathion, DDT, and phosphine (Perez-Mendoza, 1999; Pimentel et al., 2009), reducing the efficacy of these tools for the management of this stored-product pest. 42 Of course, there are limitations and problems with using behavioral control methods. For example, there is evidence that insects can become desensitized to antifeedants quite rapidly, allowing them to resume feeding in only a few hours (Bomford and Isman, 1996). There is also evidence of insects developing behavioral resistance such as reduced avoidance to DEET when previously exposed (Sfara et al., 2011). Gerold and Laarman (1964, 1967) selected for “irritability” (either hyper- or hypo-), described as either a locomotive initiator or locomotive stimulant (Miller et al., 2009), in a laboratory strain of Anopheles airoparvus (Diptera: Culicidae) using DDT-treated paper as the stimulus and found that after ten generations the resulting strains were significantly either more or less irritable than the original. As DDT acts on the insect’s nervous system, the selection for increased irritability is likely due to either a higher sensitivity of the nervous system or a decrease in penetration/detoxification resulting in smaller doses being required to stimulate irritation (Georghiou, 1972). This allows the insect to survive by terminating contact with the insecticide before a lethal dose can occur (Georghiou, 1972). However, behavioral controls can be used in conjunction with more traditional methods of control as part of an integrated pest management strategy. Naturally occurring plant compounds such as alkaloids, terpenes, and phenols have all been studied as potential antifeedant, repellent, or toxic compounds for stored-product insects (Isman, 2006; Obeng-Ofori and Reichmuth, 1997; Huang and Ho, 1998; Ho et al., 1997; Huang et al., 1999; Huang et al., 2002; Carpinella et al., 2003; Park et al., 2003). Cinnamaldehyde, a phenylpropanoid [(2E)-3-phenylprop-2-enal] (Figure 3.1), is the major constituent of cinnamon and has both toxic and antifeedant properties for S. zeamais (Huang and Ho, 1998). Eugenol, also a phenylpropanoid, is repellent to S. granaries and S. zeamais (Obeng-Ofori and Reichmuth, 1997), while safrole and isosafrole (phenylpropenes) (Figure 3.1) showed feeding deterrence to 43 Figure 3.1. Structure of several compounds that show feeding deterrence bioactivity to Sitophilus adults. 44 adult S. zeamais (Huang et al., 1999). Phenols found in plants can be oxidized by polyphenol oxidase (PPO) to quinones. Increased levels of PPOs reduce feeding by Helicoverpa armigera, cotton bollworm (Lepidoptera: Noctuidae) (Bhonwong et al., 2009). Several compounds that are likely catabolized lignin, such as dihydroxybenzenes, show promise as antifeedants to the pine weevil, Hylobius abietis, in two-choice feeding bioassays (Borg-Karlson et al., 2006). I tested a library of dialkoxybenzenes that I predicted would show similar bioactivity to the feeding-deterrent plant compounds, with hopefully minimum toxicity to the S. oryzae. These compounds are similar in structure to other naturally occurring compounds that show feeding deterrence or repellency. Therefore, by testing the structurally related compounds I also tried to determine if there were structure-activity relationships. Materials and Methods Insects All S. oryzae were obtained from laboratory colonies at the Agriculture and Agri-Food Canada labs in Winnipeg, MB, where they had been reared on whole kernels of wheat. All S. oryzae were 14-days old or younger at the start of the experiment and only individuals that visually appeared healthy (e.g. no missing limbs, actively moving, etc.) were used in the experiment. Compounds Test compounds were made as described by Paduraru et al. (2008). For a brief description please see Appendix A. DEET (N,N,-diethyl-meta-toluamide; Table 3.1) was used as the positive 45 Table 3.1. Compounds tested in a no-choice feeding bioassay using adult S. oryzae. Compound Compound name R1 R2 3c {3,3} 1,4-dipropoxybenzene Propyl Propyl 3c {4,4} 1,4-dibutoxybenzene Butyl Butyl 3c {3,6} 1-allyloxy-4-propoxybenzene Propyl Allyl 3c {6,6} 1,4-diallyloxybenzene Allyl Allyl 3c {4,6} 1-allyloxy-4-butoxybenzene Butyl Allyl 3c {n5,6} 1-allyloxy-5-pentoxybenzene n-Pentyl Allyl 3c {n5,n5} 1,4-dipentoxybenzene n-Pentyl n-Pentyl 3c {4,n5} 1-butoxy-4-pentoxybenzene Butyl n-Pentyl 3c {N2,O3} N-ethyl-4-propoxyaniline Ethyl Propyl 3c {O2,N3} N-propyl-4-ethoxyaniline Ethyl Ethyl DEET N,N,-diethyl-meta-toluamide N/A N/A 46 control as it has been shown to be repellent to S. oryzae and other species of insects (Noack and Schmidt, 1981; Khan and Wohlgemuth, 1980; Watson et al., 1997; Hou et al., 2004). Flour disks Flour disks were made using methods described by Xie et al. (1996) with some minor modifications. White flour (600 mg) was mixed using a magnetic stir bar with a total of 6 ml of liquid for a minimum of two minutes. Control flour suspensions were made with only distilled water. For each treatment a total of 1.2 ml methanol (solvent) and compound with 4.8 ml distilled water was used with enough compound to result in a concentration of 26 µmol/replicate in each flour suspension disk. Aliquots (100 µl) of the amended or control flour suspensions were pipetted onto aluminum weigh boats, partially covered with a plastic petri dish to allow airflow, and allowed to dry overnight. On the following day, five disks of one particular treatment or the no-treatment control (i.e., one treatment or control per dish) were put into individual petri dishes to be used for the feeding bioassay. The dishes were put in a growth chamber (30oC, 70% RH) and the flour disks were allowed to equilibrate for 48 hours, as fluctuations in humidity can change the weights of the flour disks (see Chapter 5). Feeding Bioassay After weighing the five flour disks, 25 S. oryzae were added to each petri dish. A plate with 25 S. oryzae but no flour disks was also made to serve as a starvation control. Flour disks were weighed (measures to 0.1mg) before the insects were added. Six replicates of each treatment were used, except for controls and solvent controls where three replicates were used. After three days, the flour disks were re-weighed to determine the amount that the insects had consumed. Mortality of the insects was checked every day after day 3 up to day 14. 47 Due to the observed rapid mortality (see results) seen at the higher dosage tested, a second dose, which was one-third the concentration for each treatment, was also tested. Flour disks were made as previously described but with a lower concentration of each compound and only 300 mg of flour used as there were only three replicates of each treatment. Experimental set-up was the same as described previously. Mortality at this dose was checked daily up to 14 days. Analysis Data were analyzed using Sigma Plot 12.5. The difference in disk weight after insect feeding (mg) was analyzed using ANOVA, followed by Tukey HSD tests when there were significant differences, or Kruskal-Wallis ANOVA on ranks followed by a Dunn’s method posthoc test, depending on whether the data met assumptions of normality and equal variance. Percent feeding reduction was determined by comparing the amount eaten by the insects to the mean amount fed on by the insects in the control plates (100%). Percent feeding reduction was analyzed using ANOVA followed by a Tukey HSD post-hoc test or Kruskal-Wallis ANOVA on ranks and followed by a Dunn’s post-hoc test. Because there was some mortality of beetles during the first three days, the difference in disk weight and percent feeding reduction were adjusted to account for the loss of insects by calculating the number of beetles feeding on the disks over the three days and determining the feeding per beetle-day. For the high dose, as there was uncertainty on which day the insects died, it was assumed they died on day two. These data were analyzed in the same manner as the disk weight and percent feeding reduction. The median lethal time (LT50) was also calculated for each treatment using a Kaplan-Meier log-rank survival analysis and the Holm-Sidak method of 48 multiple comparisons. To determine if there was a relationship between the amount of flour disk consumed and the insect survival (# insects alive at 14 days/# insects at start), a Pearson’s product moment correlation was used for both doses. Results At the higher concentration (26 µmol/replicate), DEET caused the highest feeding reduction of the compounds tested but this was not the case at the lower concentration (8.7 µmol/replicate) (Table 3.2, Table 3.3). All compounds at the high concentration caused some level of feeding reduction when compared to the no-treatment control (Table 3.2) (amount consumed (mg): H11 = 56.30, P < 0.001). At the low concentration not all compounds caused feeding reduction (e.g. 3c{3,3}, 3c{3,6}, 3c{4,6}, 3c{6,6}); however, there were several test compounds that caused statistically significant feeding reduction by S. oryzae (Table 3.3). Even when the mortality that occurred during the first three days was accounted for (feeding reduction per beetle day: F12,26 = 71.42, P < 0.001), there were still significant reductions in feeding on the treated disks compared to the control disks. The most effective feeding deterrent tested was 3c{N2,O3}. In addition, DEET, 3c{5,5}, 3c{4,5}, and 3c{4,4} all reduced feeding by over 50%. There were significant differences in the toxicity of the treatments in both the high and low doses (High: χ213 = 1130.2, P < 0.001; Low: χ213 = 1210.5, P < 0.001). At the higher dose, all of the compounds caused mortality that was higher than the controls and some, such as 3c{3,6}, 3c{4,4}, 3c{4,5}, 3c{5,5}, 3c{5,6}, 3c{N2,O3}, 3c{O2,N2}, had a lower LT50 than the starving control. At the lower dose, several of the compounds had the same LT50 as the control, and only one of the compounds, 3c{4,5}, showed a lower LT50 than the starvation controls. 49 Table 3.2. Feeding deterrence and toxicity of feeding bioassays by S. oryzae at the high dose of 26 µg/replicate (6 replicates/treatment) after three days. Significant differences in columns for amount consumed at day three and percent feeding reduction (P < 0.05) are represented by different letters. Measurements per beetle day were adjusted to account for the number of beetles feeding on each plate due to mortality during the first three days. Mean percent feeding for the control plates was set at 100%. Treatment Amount consumed at day three (mg) (± SE) % Feeding reduction (± SE) Amount consumed at day three (mg/beetle day) (± SE) % Feeding reduction by beetle day (± SE) Survival (# insects alive at 14 days/# insects at start) (± SE) LT50 (days) (± SE) 3c{3,3} 20.0 (± 1.5)a 55.4 (± 4.1)a 0.27 (± 0.02)b 54.5 (± 3.7)b 0.27 (± 0.02) 9.7 (± 0.3)b 3c{3,6} 6.2 (± 0.7)ab 17.2 (± 2.0)ab 0.14 (± 0.01)bc 27.7 (± 2.6)bcd 0.06 (± 0.03) 4.0 (± 0.2)ef 3c{4,4} 4.4 (± 0.2)b 12.1 (± 0.5)b 0.07 (± 0.00) c 14.8 (± 0.6)cde 0.00 (± 0.00) 3.7 (± 0.1)f 3c{4,5} 6.8 (± 1.0)ab 18.8 (± 2.7)ab 0.11 (± 0.02) c 23.0 (± 3.4)cde 0.01 (± 0.01) 3.7 (± 0.1)f 3c{4,6} 12.9 (± 3.8)ab 35.7 (± 10.5)ab 0.19 (± 0.05)bc 38.9 (± 9.6)bc 0.29 (± 0.13) 7.6 (± 7.6)bc 3c{5,5} 5.6 (± 0.2)ab 15.5 (± 0.4)ab 0.11 (± 0.00)c 21.5 (± 0.6)cde 0.00 (± 0.01) 3.6 (± 0.1)f 3c{5,6} 7.0 (± 0.9)ab 19.3 (± 2.5)ab 0.11 (± 0.01)c 21.9 (± 2.0)cde 0.02 (± 0.01) 4.5 (± 0.2)e 3c{6,6} 12.5 (± 4.8)ab 34.7 (± 13.4)ab 0.25 (± 0.08)b 50.4 (± 16.3)b 0.04 (± 0.03) 4.4 (± 0.3)def 3c{N2,O3} 3.9 (± 0.2)b 10.9 (± 0.6)b 0.12 (± 0.01)c 24.0 (± 2.1)cde 0.00 (± 0.00) 3.0 (± 0.0)g 3c{O2,N2} 4.3 (± 0.1)b 12.0 (± 0.4)b 0.18 (± 0.01) bc 36.7 (± 2.5)bc 0.00 (± 0.00) 3.0 (± 0.0)g DEET 3.6 (± 0.3)b 10.0 (± 0.8)b 0.05 (± 0.00)d 10.1 (± 0.8)e 0.05 (± 0.03) 5.5 (± 0.2)cd Control 37.0 (± 1.0)a 100.0a 0.50 (± 0.01)a 100.0a 0.83 (± 0.09) 13.7 (± 0.1)a Starving* - - - - - 6.4 (± 0.3)cd * One plate of S. oryzae with no flour disks used in the experiment. 50 Table 3.3. Feeding deterrence and toxicity results for feeding bioassays by S. oryzae at the low dose of 8.7 µg/replicate (n = 3) after three days. Significant differences in columns (P < 0.05) are represented by different letters. Measurements per beetle day were adjusted to account for the number of beetles feeding on each plate due to mortality during the first three days. Mean percent feeding for the controls was set at 100%. Treatment Amount consumed at day three (mg) (± SE) % Feeding reduction (± SE) Amount consumed at day three (mg/beetle day) (± SE) % Feeding reduction per Survival (# insects alive at beetle day (± SE) 14 days/# insects at start) (± SE) LT50 (days) (± SE) 3c{3,3} 36.3 (± 0.8)ab 97.5 (± 2.2)ab 0.5 (± 0.01)a 97.5 (± 2.2)ab 1.0 (± 0.03) 14.0 (± 0.0)e 3c{3,6} 37.5 (± 1.1)ab 100.7 (± 2.9)a 0.5 (± 0.01)a 100.7 (± 2.9)a 1.0 (± 0.0) 14.0 (± 0.0)e 3c{4,4} 14.0 (± 1.1)e 37.6 (± 2.9)ef 0.2 (± 0.01)de 38.4 (± 2.5)ef 0.08 (± 0.05) 7.6 (± 0.4)c 3c{4,5} 7.2 (± 0.8)ef 19.3 (± 2.2)fg 0.1 (± 0.01)ef 22.2 (± 2.3)fg 0.0 (± 0.0) 4.6 (± 0.2)a 3c{4,6} 29.6 (± 1.2)bc 79.6 (± 3.3)bc 0.4 (± 0.01)ab 80.6 (± 2.3)bc 1.0 (± 0.01) 13.8 (± 0.2)e 3c{5,5} 7.3 (± 1.8)ef 19.6 (± 4.8)fg 0.1 (± 0.03)f 20.9 (± 5.3)fg 0.0 (± 0.0) 5.2 (± 0.2)ab 3c{5,6} 19.9 (± 2.7)d 53.6 (± 7.1)de 0.3 (± 0.03)cd 54.2 (± 6.8)de 0.6 (± 0.04) 11.3 (± 0.5)d 3c{6,6} 38.3 (± 0.5)a 103.0 (± 1.4)a 0.5 (± 0.01)a 103.0 (± 1.4)a 1.0 (± 0.01) 14.0 (± 0.0)e 3c{N2,O3} 2.1 (± 0.5)f 5.7 (± 1.4)g 0.03 (± 0.01)g 6.5 (± 1.3)g 0.07 (± 0.05) 5.1 (± 0.3)ab 3c{O2,N2} 24.2 (± 2.5)cd 65.0 (± 6.8)cd 0.3 (± 0.03)bc 65.0 (± 6.8)cd 1.0 (± 0.01) 14.0 (± 0.0)e DEET 11.1 (± 1.7)e 29.7 (± 4.6)f 0.1 (± 0.02)ef 29.7 (± 4.6)f 0.9 (± 0.0) 13.5 (± 0.2)e Control 37.2 (± 2.3)ab 100.0a 0.5 (± 0.03)a 100.0ab 1.0 (± 0.01) 13.9 (± 0.1)e Control MeOH 35.2 (± 1.6)ab 94.7 (± 4.4)ab 0.5 (± 0.02)a 95.9 (± 3.3)ab 1.0 (± 0.02) 13.6 (± 0.2)e Starving* - - - - - 6.1 (± 0.2)bc * One plate of S. oryzae with no flour disks used in experiment. 51 There was a significant positive relationship between the amount of flour disk consumed after three days (mg/beetle day) and the mortality after 14 days for both doses (High: r = 0.80, P < 0.001, n = 72; Low: r = 0.82, P < 0.001, n = 39). As feeding deterrence increased, the mortality also increased (Figure 3.2). Discussion There were significant differences among treatments, and treatments and controls, in both toxicity and feeding reduction. These differences were found at both concentrations tested. In addition, there appears to be some structure-activity relationships between the compound structure and the feeding deterrence and toxicity observed. Based on both the high and low doses tested, the compounds with the smallest substitutions (propyl and allyl) showed lower toxicity but were also less effective at reducing feeding. Compounds substituted with longer chains (pentyl and butyl) caused higher feeding deterrence but also exhibited lower LT50. There were significant correlations between the lethal effect of the compounds and feeding deterrent effects for both doses. Interestingly, even substituting one pentyl with an allyl (3c{5,5} vs 3c{5,6}) drastically reduced the percent feeding reduction per beetle-day at both doses (Table 3.2). The most effective compound, 3c{N2,O3} that caused a feeding reduction, particularly at the low dose, does have structural similarities to DEET, which is a known repellent to some storedproduct insects (Khan and Wohlgemuth, 1980, Hou et al., 2004). However, 3c{N2,O3} appeared to be more toxic to S. oryzae than DEET at both doses. 52 Figure 3.2. Relationship between mean survival and mean amount of disk consumed by S. oryzae (±SE). Disks were treated with one of ten dialkoxybenzenes at either 5 µg/replicate (high dose) or 1.7 µg/replicate (low dose). Correlations were significant for both doses (P < 0.05). 53 3c{3,6} is a feeding deterrent to Trichoplusia ni, the cabbage looper, (Cameron et al., 2014) and at the high dose I found it deterred some feeding by S. oryzae but not at the low dose. A dose-dependent response was also observed for T. ni (Cameron et al., 2014). However, 3c{3,6} was not the most effective feeding deterrent I tested. At the highest dose tested, DEET was an effective feeding deterrent but it was not significantly more effective than several others such as 3c{4,4}, 3c{4,5}, 3c{5,5}, 3c{5,6}, and 3c{N2,O3}. However, all of the other test compounds showed significantly lower LT50 than the starving control and the control flour disks. This increased toxicity indicates that at the higher concentration these compounds may not be effective at controlling behavior without killing the insects as well. For both doses I observed a strong positive correlation between feeding deterrence and mortality. At the low concentration tested, the most effective feeding deterrent was 3c{N2,O3}, while DEET, 3c{5,5}, 3c{4,5}, and 3c{4,4} all reduced feeding by over 50%. None of these showed a lower LT50 than the starving control except for 3c{4,5}. That the S. oryzae did not feed despite the fact that the insects should have been highly motivated because they were starving indicates that these are strong feeding deterrents. The concept of motivation affecting behavior has been reviewed thoroughly by Dethier (1976). In brief, the internal state of the insect will affect its behavior, and responses to food will be stronger if they have been deprived of it (Dethier, 1976, Bowdan and Dethier, 1986). Feeding was measured after three days which means the insects were not starving during that time (starving LT50 = 6.4 ± 0.3; 6.1 ± 0.2 days) but as time progressed, the insects would have been more and more motivated to feed (as they starved). Therefore, the insects could have then fed and died due to toxicity. Alternately, the feeding deterrence stimulus could have been so strong 54 they died of starvation, which has been documented using azadirachtin and other species of insects (Mordue (Luntz), 2007; Simmonds et al., 1990). I did not observe any indications of neurotoxicity (e.g., twitching) among test insects, but that does not eliminate this as a potential explanation for the mortality observed. Reduced growth of larval Helicoverpa armigera fed a diet containing salannin-type compounds did not differ from starvation controls, indicating strong antifeedant activity without toxicity (Koul et al., 2004). As the compounds showed lethality at the higher dose, it is likely that eventually the motivation to feed became so strong that they ate and then died from a toxic effect even though the insects avoided the disks at the beginning. The feeding deterrent effects could still be used as a control method particularly, if an attractive alternative is offered, thus reducing the motivation to feed where the feeding deterrent compound exists. As an example, the push-pull strategy uses an unattractive stimulus to “push” the insect away in addition to using an attractive stimulus (pull) to move populations away from a protected resource (Pyke et al., 1987, Cook et al., 2007). One of the compounds I tested showing feeding deterrent properties (e.g., 3c{3,6}) could be used as the “push” stimulus to successfully protect a resource. If the insects do not like to feed on the treated material, it could be used in the same manner as extract from the Neem tree was by Pyke et al. (1987), and the aggregation pheromone of S. oryzae could be used as the pull to draw the insects away from the desirable resource into an area where they cannot do damage or where they can be killed. Reducing the concentration of the test compounds by one-third reduced the toxicity of the treatments and, in general, also reduced the efficacy of any feeding deterrence (Table 3.3). At the lower dose, DEET was not the most effective feeding deterrent tested – 3c{N2,O3} was. 55 Surprisingly, at the lower dose, the amount of flour disk fed on by the beetles (adjusted for mortality of the insects during the first three days) was actually less than at the higher dose (0.03 ± 0.01 vs. 0.12 ± 0.01 mg). It is difficult to explain why a lower dose would result in greater feeding deterence. Studies using alkaloid treated leaf disks showed feeding deterence for nine alkaloids by larval gypsy moths showed a positive correlation between feeding deterrence and concentrations (Shields et al., 2008) as did feeding deterrence of S. zeamais by cinnamaldehyde (Huang and Ho, 1998). Further, at the lower dose the LT50 was not significantly different than observed in the beetles that were starved during the test period, although it was significantly lower than observed with DEET. This indicates that 3c{N2,O3} could prove to be an effective feeding deterrent at relatively low doses for beetles but it does result in mortality. It could still be effective in controlling stored-product insects particularly if used in a push-pull strategy as discussed above. Why the compounds I tested acted as feeding deterrents is more difficult to determine. Because I accounted for mortality in determining the percent feeding reduction, not all of the feeding reduction can be due to an immediate lethality. Some of the compounds tested are volatile enough that the insects could be deterred from feeding due to olfactory detection (Ebrahimi et al., 2013). Sitophilus oryzae has been shown to be attracted to several grain volatiles in laboratory assays, as well as to their aggregation pheromone (Phillips et al., 1993), suggesting adult rice weevils are capable of making behavioral choices based on olfaction. The compounds could be acting as an antifeedant by affecting peripheral sensilla (gustation) (Isman et al., 2006). For example, the ratio of deterrent and phagostimulatory compounds is believed to determine host selection by phytophagous insects (Chapman, 2003). 56 The compounds could also be deterring feeding based on a systemic effect on the insect postingestion, or on a sublethal toxic effect. For example, larvae of the tobacco hornworm (Manduca sexta) rejected diet containing a feeding deterrent even after mouthpart chemoreceptors were ablated, suggesting that the food rejection was due to a postingestive effect rather than gustation (Glendinning, 1996). Behavioral effects, including feeding deterrence and repellency, have been documented due to sublethal levels of insecticides (Haynes, 1988, Desneux et al., 2007). The sublethal toxic effects on the parisitoid Microplitis croceipes after feeding on insecticide-treated cotton nectar resulted in a signficantly reduced ability to forage (Stapel et al., 2000). While I could determine some general information about the structure-activity relationship, future work could further elucidate this by focusing on the mechanism by which the compounds are interacting with insects, resulting in behavioral bioactivity. For example, how these compounds might bind with receptors or odorant binding proteins could provide more information on how the behavioral response is elicited and potentially show new avenues for refining the structure to make the compounds more bioactive. The binding affinity of several of these types of compounds has been determined for pheromone binding proteins of gypsy moth (Lymantria dispar) (Paduraru et al., 2008). There would need to be extensive additional testing before these compounds could be used in an operational setting to determine characteristics such as vertebrate toxicity and persistance just to name a few. In addition, futher research could help explain how these compounds elicit the response they do. However, based on the feeding bioassay, some of the tested compounds could have potential as a feeding deterrent to S. oryzae in certain doses and could be useful as a control tactic in a management strategy. 57 Chapter 4. Feeding choices of five species of stored-product insects to dialkoxybenzenes Abstract I tested a library of dialkyoxybenzenes using a no-choice feeding bioassay to determine if the compounds have feeding deterrent activity or toxicity to five species of stored-product insects: Sitophilus oryzae, S. zeamais, Tribolium castaneum, T. confusum, and Rhyzopertha dominica. Insects were fed wheat flour disks treated with two doses of eight test compounds. Differences in disk weight were measured after three days of feeding and mortality was assessed daily for 14 days after initial exposure. There were significant differences in both feeding and mortality between the five species including the closely related species. The primary pests (S. oryzae, S. zeamais, R. dominica) did show more sensitivity to the test compounds in general compared with the secondary pests (T. castaneum and T. confusum), which may be explained by the difference in the range of materials they can use (narrower vs. broader respectivly). I found two compounds that caused reduced feeding without increasing mortality (3c{2,2} and 3c{4,4}), which may have the potential to be developed into new feeding deterrent treatments in operational settings. Introduction Insects that infest stored products cause a wide range of damage including physical loss and spoilage, contamination through insect parts or waste (e.g., Tribolium castaneum contaminates flour with benzoquinones [Markarian et al., 1978, Hodges et al., 1996]). There are also costs and health risks associated with control measures such as the application of insecticides, to list just a few (Rees, 2004). It is estimated that grain and food products lost to 58 pests, including insects and mites, account for approximately 10-15% of the annual food supply worldwide (Rajendran, 2002). Thus, there is substantial interest in the development of new and effective tactics and strategies to protect stored products from insect damage. Standard control strategies include the use of insecticides (White and Leesch, 1996, Arthur and Subramanyam, 2012, Phillips et al., 2012), while alternatives include biological control (parasitoids, natural predators, etc.) (Brower et al., 1996; Schöller, 2010; Schöller et al., 2006), physical control (includes control by temperature) (Fields and Muir, 1996), mating disruption (Phillips, 1997; Campos and Phillips, 2014; Burks and Kuenen, 2012) and behavioral control (Foster and Harris, 1997). The behavioral manipulation of a pest uses three principal elements: the identification of a behavior of a pest, an ability to manipulate the behavior, and a way to use the behavioral manipulation to protect a resource (Foster and Harris, 1997). The ability to manipulate behavior, including the use of push-pull strategies (Pyke et al., 1987; Cook et al., 2007), depends on several factors. These include accessibility of the stimulus, reproducibility of the stimulus, how well the stimulus can be controlled, specificity of the stimulus which increases the likelihood that it can manipulate the behavior, and how practical the use of the stimulus is which can include cost and other hazards that may exist in its use (Foster and Harris, 1997). In some ways, the criteria that contribute to the efficacy of an insecticide also apply to behavioral modification, such as being inexpensive, having low non-target toxicity (such as to humans or to pollinators), being easy to handle, etc. (Isman, 2002; White and Leesch, 2006). For example, if the stimulus has high toxicity to other organisms then it may not be a useful tool to use in pest management strategies even if it is successful at manipulation of pest behavior. 59 Some naturally occurring compounds reduce feeding, have repellent effects, or increase mortality in stored-product insects (Figure 4.1). Several sesquiterpenes, including chlorojanerin, as well as syringin (a phenylpropanoid glycoside) are antifeedants to stored-product insects including Sitophilus granarius and Tribolium confusum (Nawrot et al., 1986; Cis et al., 2006). Some triterpenes show antifeedant activities against S. oryzae (Omar et al., 2007). Turmerone [2methyl-6-(4-methyl-1,4-cyclohexadien-1-yl)-2-hepten-4-one] and ar-turmerone [2-methyl-6-(4methylphenyl)-2-hepten-4-one], which are sesquiterpenes isolated from turmeric, repelled adult T. castaneum (Su et al., 1982). Cinnamaldehyde [(2E)-3-phenylprop-2-enal] showed significant antifeedant effects against adult S. zeamais but not against adult T. castaneum at the rates tested (Huang and Ho, 1998). A study on acaricidal properties of some monoterpenes found that molecules possessing a free alcohol or phenol group showed the most activity (Perrucci et al., 1995). Cinnamaldehyde and eugenol, both phenylpropaniods, have toxic, antifeedant, and repellent properties to Sitophilus sp. (Obeng-Ofori and Reichmuth, 1997; Huang and Ho, 1998; Huang et al., 2002). Benzyl alcohol and 2-phenylethanol show feeding deterrence for the pine weevil (Hylobius abietis) (Eriksson et al., 2008). 60 Figure 4.1. Examples of some naturally occurring compounds that show feeding deterrence against some stored-product insects. Syringin Chlorojanerin Turmerone Eugenol ar-Turmerone Benzyl alcohol Cinnamaldehyde 2-Phenylethanol 61 I tested a library of dialkoxybenzenes, which are phenol derivatives (Paduraru et al., 2008) (Table 4.1), for feeding deterrence and toxicity against a variety of stored-product pests. The structures of these compounds were similar to some of the naturally occurring compounds that exhibit feeding deterrence and toxic effects on stored-product insects. I predicted that these compounds would therefore also show behavioral activity, but as I was using a library of related structures it was also possible to determine if there were structure-activity relationships (Akhtar et al., 2007) which would be beneficial as a guide for future research into compounds that have potential for use as behavioral control agents. I conducted feeding bioassays to determine feeding and mortality properties of phenolderived compounds using five important species of stored-product insects: Tribolium castaneum, T. confusum, Sitophilus oryzae, S. zeamais, and Rhyzopertha dominica. Tribolium castaneum, the red flour beetle, and T. confusum, the confused flour beetle, (Coleoptera: Tenebrionidae), both feed on grain that has already been damaged or milled, making them secondary pests. Sitophilus oryzae, the rice weevil, and S. zeamais, the maize weevil, (Coleoptera: Curculionidae), are capable of damaging whole kernels, making them primary pests. Rhyzopertha dominica, the lesser grain borer, (Coleoptera: Bostrichidae), is also capable of damaging whole kernels. All five of these species have aggregation pheromones (Phillips et al., 2000) and are attracted to food volatiles (Edde, 2012; Ryan and O’Ceallachain, 1976; Faustini et al., 1981; Phillips et al., 1985; Dowdy et al., 1993; Phillips et al., 1993; Trematerra et al., 2000; Ukeh et al., 2009), indicating they are capable of making behavioral choices based on chemical signals. They are also considered long lived insects (Hagstrum and Subramanyam, 2006). For example, T. confusum 62 can live over 730 days, S. oryzae can live 7-8 months (Sinha and Watters, 1985), and R. dominica adults live on average over 100 days (Edde, 2012). While all the species tested are Coleopterans, the three genera are not closely related (Hunt et al., 2007). For this bioassay, I used adults of each species as they are the most mobile life-stage and the stage that makes the host-selection decision. I predicted that the species that are closely related would show similar antifeedant and toxic responses to the test compounds. I also predicted that Sitophilus oryzae, S. zeamais, and R. dominica would show more behavioral sensitivity to the experimental treatments as they are primary pests, while the two species of Tribolium would show lower behavioral bioactivity to the compounds. Primary pests tend to use a narrower range of materials than do secondary pests (Rees, 2004). Sitophilus spp. are often considered cereal seed specialists whereas Tribolium spp. is more of a generalist. Therefore, I predicted that the greater host range occupied by the Tribolium spp. would result in a better ability to successfully survive and feed on the various compounds. Materials and Methods Insects Adult insects used were from laboratory colonies reared at the Agriculture and Agri-Food Canada laboratory in Winnipeg, Manitoba. Tribolium castaneum and T. confusum were reared on whole wheat flour mixed with 5% b.w. brewer’s yeast. Sitophilus oryzae, S. zeamais, and R. dominica were reared on whole kernels of wheat. Only insects that appeared visually healthy (e.g., moving, all body parts visibly present) were used in the experiments. To determine the average weight of an individual insect for each of the five species tested, 100 insects from the 63 colony used in the experiment were weighed in ten replicates of ten randomly chosen insects (accuracy of 0.001g). Data were analyzed using a Kruskal-Wallis one way analysis of variance on ranks as data failed to meet requirements of normality, followed by a Dunn’s method posthoc. Experimental compounds The dialkoxybenzenes used for this experiment (Table 4.1) were synthesized as described by Paduraru et al., (2008) (Appendix A) and selected based on the activity seen in the previous bioassays (Chapters 2 and 3). Briefly, the compounds that showed the most bioactivity against T. castaneum were para-substituted and were substituted with smaller/medium length chains and many of these compounds also showed bioactivity with S. oryzae. Flour disks Flour disks were made using slightly modified methods described by Xie et al. (1996). White flour (2.5 g) was mixed with distilled water (12.5 mL) using a magnetic stir bar for a minimum of two minutes. Control flour disks were made with only distilled water or distilled water and the carrier solution, methanol (HPLC grade). Treatment disks were made using individual test compounds dissolved in methanol at two concentrations (1 mL total liquid per 200 mg flour). The high concentration was equivalent to 26 µmol of test compound per replicate (approximately 5 mg compound per replicate) and the low concentration was 1/3 the high dose amount (8.7 µmol per replicate). Every treatment was mixed in one batch to ensure that each species was feeding on identical flour disks. Aliquots (100 µL) were pipetted onto aluminum 64 Table 4.1. Abbreviations used to name each experimental compound used in the feeding bioassay. The two functional groups attached to the basic molecule structure and the compound name is listed. Treatment R1 R2 Compound 3c{2,2} ethyl ethyl 1,4-diethoxybenzene 3c{3,3} propyl propyl 1,4-dipropoxybenzene 3c{4,4} butyl butyl 1,4-dibutoxybenzene 3c{3,6} propyl allyl 1-(allyloxy)-4-propoxybenzene 3c{n5,6} n-pentyl allyl 1-(allyloxy)-4-pentoxynbenzene 3c{n5,n5} n-pentyl n-pentyl. 1,4-bis(pentyloxy)benzene 3c{4,n5} butyl pentyl 1-butoxy-4-(pentyloxy)benzene DEET N/A NA N,N-diethyl-3-methylbenzamide 65 weigh boats and then allowed to dry overnight, partially covered with a petri dish. The following day the dry disks were put into new petri dishes and placed in a growth chamber (30oC, 70% RH) for 24 hours to allow the flour disks to equilibrate (as changes in humidity can affect weight due to moisture content). Feeding bioassay Twenty-five insects of the same species were added to each petri dish containing five flour disks (same treatment). The flour disks were weighed to an accuracy of 0.001g before the insects were added and again after three days to determine the amount of flour eaten. Mortality of the insects was checked every day after the flour disks were weighed until the experiment had run for 14 days. Five replicates of each treatment for each dose were tested for each of the five species. Some treatments had only four replicates as there were occasionally not enough successful flour disks made (see Results). It was not logistically possible to test all compounds for all species on one day; therefore the experiments were started over four days with the same treatment for all five species started on the same day. Analysis All data were analyzed using SigmaPlot 12.5. Unless otherwise indicated the data were analyzed using either ANOVA followed by a Tukey HSD pairwise comparison if there were significant differences, or a Kruskal-Wallis one-way analysis of variance on ranks if data failed to meet the assumptions of normality (Shapiro-Wilk normality test) or equal variance, followed by a Dunn’s method post-hoc analysis if there were signficant differences. Because some mortality occurred during the first three days, before the disk weight was measured to determine 66 the amount the beetles fed on, the number of beetles that fed for each day (beetle-day) was calculated and the difference between the disk weights was divided by that amount, making the assumption that beetles died on the second day to try and approximate the amount of feeding by each beetle. The amount eaten relative to the controls by the beetles was calculated by setting the average amount fed on by the beetles on control disks to 100%. Survival was calculated by dividing the number of insects alive at the end of the experiment (14 days) by the initial number of insects feeding. The median lethal time (LT50) was calculated using a log-rank Kaplan-Meier survival analysis with significant differences between treatments determined using a Holm-Sidak method pairwise multiple comparison (P = 0.05). There were several treatments where there was no mortality and no LT50 could be calculated. Because the species used for the no-choice feeding assay were different sizes (Table 4.2), the difference in disk weight per beetle-day after three days was also divided by the average weight of the insect species to standardize the measure of feeding. A Pearson correlation was used to assess the potential relationship between survival and the amount fed on per beetle-day. The hydrophobic/hydrophilic nature of some of the test compounds was determined by calculating the octanol-water partition coefficients (Log Kow) (Shaima Kammoonah, SFU, personal communication) and a Pearson correlation was used to asses if there was a relationship between the Log Kow and both the percent feeding reduction per beetle-day and the mortality for the high dose as the hydrophobicity could be related to the bioactivity of the compounds. 67 Results There were significant differences between the mean weights of the five species of insects tested (H4 = 46.08, P < 0.001) (Table 4.2). Sitophilus zeamais was the heaviest insect but also had the greatest variance of the species tested. Rhyzopertha dominica was the lightest insect but was not significantly different than S. oryzae (Table 4.2). There were significant differences (P < 0.05) between treatments for all five species used in the feeding bioassays (Table 4.3-4.7). Sitophilus oryzae was deterred from feeding by three compounds at the high dose compared to the controls, based on the mean amount consumed per beetle-day (3c{4,4}, 3c{4,n5}, DEET) (H9 = 44.14, P < 0.001), but all the compounds tested except for one (3c{2,2}) resulted in a lower LT50 than the controls (χ210 = 1108.92, P < 0.001) (Table 4.3). At the lower dose, the compounds showed a reduced efficacy in deterring feeding by S. oryzae except for 3c{4,n5} and DEET (H9 = 43075, P < 0.001), but several compounds including 3c{4,n5} and DEET still had lower LT50 than the controls (Table 4.3) (χ210 = 1190.00, P < 0.001). There was a significant correlation between amount ingested of all the compounds tested (mg/beetle-day) and the mortality for both doses tested (High: r = 0.78, P < 0.001, n = 48; Low: r = 0.62, P < 0.001, n = 49) (Figure 4.2). Sitophilus zeamais was deterred from feeding by several compounds at the high dose (3c{3,3}, 3c{4,4}, 3c{3,6}, 3c{n5,6}, 3c{n5,n5}, 3c{4,n5}, and DEET) compared to the controls (F9,39 = 150.51, P < 0.001), and all compounds tested had a lower LT50 than the controls except for 3c{2,2} (χ210 = 919.73, P < 0.001) (Table 4.4). 68 Table 4.2. Mean weight of the five Coleopteran species (±SE) used in the feeding bioassays. Data were analyzed using a Kruskal-Wallis one way analysis of variance on ranks, followed by a Dunn’s method post-hoc analysis. Significant differences are represented by different letters (P < 0.05, n = 10). Species Weight of a single insect (mg) Rhyzopertha dominica 1.3 (± 0.01)d Sitophilus oryzae 1.6 (± 0.03)cd Tribolium castaneum 2.1 (± 0.02)bc Tribolium confusum 2.8 (± 0.02)ab Sitophilus zeamais 4.8 (± 2.3)a 69 At the lower dose, fewer of the compounds reduced feeding when compared to the controls (3c{n5,6} and DEET) (H9 = 42.51, P < 0.001), but again resulted in a lower LT50 than the controls (χ210 = 636.70, P < 0.001) (Table 4.4). There was a significant correlation between amount fed (mg/beetle-day) and the mortality for both doses tested (High: r = 0.72, P < 0.001, n = 49; Low: r = 0.53, P < 0.001, n = 49) (Figure 4.3). Tribolium castaneum was deterred from feeding by 3c{4,4} at the high and low dose compared to the controls (High: H9 = 38.68, P < 0.001; Low: H9 = 43.00, P < 0.001). At the high dose, only 3c{3,6} showed a lower LT50 than the controls (χ210 = 514.99, P < 0.001). In the low dose bioassays, several compounds reduced feeding by T. castaneum including 3c{4,n5}, 3c{n5,6}, 3c{n5,n5}, and DEET (Table 4.5). There was no significant correlation between feeding and mortality for either dose (High: r = 0.12, P = 0.42, n = 48; Low: r = -0.03, P = 0.82, n = 50) (Figure 4.4). Tribolium confusum showed no reduction in feeding (per beetle-day) for any of the treatments compared to the control, although there were significant differences between some of the treatments (High: H9 = 31.59, P < 0.001; Low: H9 = 37.89, P < 0.001) There also were no differences between the LT50 for any of the treatments except for the starvation control (High: χ210 = 286.34, P < 0.001; Low: χ210 = 553.35, P < 0.001) (Table 4.6). There was also no significant correlation between feeding and mortality for either dose (High: r = 0.23, P = 0.11, n = 49; Low: r = 0.11, P = 0.46, n = 50) (Figure 4.5). Rhyzopertha dominica was deterred from feeding by 3c{4,n5} and 3c{2,2} at the high dose and 3c{4,4} and 3c{4,n5} at the low dose (High: H9 = 42.79, P < 0.001; Low: H9 = 43.81, P < 0.001) (Table 4.7). These compounds also had a lower LT50 than the controls except for 70 Table 4.3. Feeding and mortality (mean ± SE) of S. oryzae in no-choice feeding bioassay on flour disks with different compounds. High dose flour disks were treated with 26 µmol/replicate and low dose flour disks were treated with 8.67 µmol/replicate. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Compound N, Food H/L consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetle-day (%) – High* Feeding/ beetle-day (%) – Low* Survival at 14 days – High Survival at 14 days – Low LT50 – LT50 – High Low (d) (d) 3c{2,2} 4/5 31.6 (1.6)a 33.2 (±1.6)a 101.8 (±5.1)a 107.3 (±5.2)a 0.42 (±0.01)ab 0.45 (±0.03)a 100.0 (±3.8)a 108.0 (±6.9)a 0.6 (±0.2) 0.5 (±0.2) 12.4 (0.3)a 13.1 (0.2)b 3c{3,3} 5/5 9.2 (1.2)abc 24.3 (±1.0)abc 31.0 (±4.0)abc 81.9 (±3.4)abc 0.13 (±0.01)abc 0.33 (±0.02)abc 30.1 (±3.4)abc 78.8 (±4.1)abc 0.0 (±0.0) 0.7 (±0.1) 5.8 (0.2)bc 12.9 (0.2)ab 3c{4,4} 5/5 0.7 (0.9)c 2.3 (±0.7)bc 2.3 (±3.2)c 7.7 (±2.4)c 0.01 (±0.01)c 0.03 (±0.01)c 2.3 (±3.4)bc 8.3 (±2.5)bc 0.0 (±0.0) 0.0 (±0.0) 4.2 (0.1)e 4.4 (0.1)e 3c{3,6} 5/5 3.0 (0.1)abc 11.9 (±2.8)abc 9.8 (±0.4)abc 38.2 (±9.0)abc 0.07 (±0.00)abc 0.17 (±0.04)abc 15.8 (±0.7)abc 39.7 (±9.2)abc 0.0 (±0.0) 0.7 (±0.1) 3.7 (0.1)f 11.5 (0.4)ab 3c{n5,6} 5/5 3.9 (0.3)abc 3.6 (±0.7)bc 12.7 (±0.9)abc 11.7 (±2.3)bc 0.07 (±0.01)abc 0.06 (±0.01)bc 15.8 (±1.2)abc 13.7 (±2.7)abc 0.0 (±0.0) 0.0 (±0.0) 4.2 (0.1)de 4.7 (0.1)e 3c{n5,n5} 4/5 2.3 (0.3)abc 6.7 (±0.6)abc 17.7 (±1.0)abc 22.8 (±2.1)abc 0.08 (±0.00)abc 0.09 (±0.01)abc 19.1 (±1.1)abc 22.2 (±1.9)abc 0.0 (±0.0) 0.0 (±0.0) 4.9 (0.1)c 5.8 (0.1)df 3c{4,n5} 5/4 -0.9 (0.3)c 0.3 (±0.2)c -3.0 (±0.8)c 1.0 (±0.8)c -0.02 (±0.00)c 0.00 (±0.00)c 2.3 (±1.0)c 1.2 (±0.8)bc 0.0 (±0.0) 0.0 (±0.0) 4.7 (0.2)cd 5.1 (0.2)df DEET 5/5 2.7 (0.1)bc 8.4 (±3.9)abc 9.2 (±0.4)bc 28.3 0.04 (±13.1)abc (±0.00)bc 0.11 (±0.05)abc 9.9 (±0.4)bc 0.4 (±0.2)c 0.0 (±0.0) 0.2 (±0.1) 4.2 (0.1)e 7.1 (0.3)cf Control 5 31.0 (0.8)a 31.0 (±0.8)a 100ab 100ab 0.42 (±0.01)a 0.42 (±0.01)ab 100a 100a 0.5 (±0.2) 0.5 (±0.2) 13.4 (0.1)a 13.4 (0.1)ab Methanol 5 27.8 (1.4)ab 27.8 (±1.4)ab 93.8 (±4.7)ab 93.8 (±4.7)ab 0.38 (±0.01)ab 0.38 (±0.01)ab 91.7 (±2.6)ab 91.7 (±2.6)ab 0.7 (±0.2) 0.7 (±0.2) 13.4 (0.2)a 13.4 (0.2)a Starvation 1 - - - - - - - - 0.0 0.0 6.1 (0.3)b 6.1 (0.3)f *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 71 Figure 4.2. Correlation of the percent feeding by S. oryzae after three days and the average percent survival after 14 days (±SE) for all the compounds tested. There was a significant correlation for both doses tested. 72 Table 4.4. Feeding and mortality (mean ± SE) of S. zeamais in no-choice feeding bioassay on flour disks with different compounds. High dose flour disks were treated with 26 µmol/replicate and low dose flour disks were treated with 8.67 µmol/replicate. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Compound N, H/L Food consumed (mg) – High Food consumed (mg) – Low 3c {2,2} 4/5 47.4 (2.2)a 3c {3,3} 5/5 3c {4,4} Feeding (%) – Low* Food consumed (mg/beetle -day) – High Food consumed (mg/beetle -day) – Low Feeding/ beetle-day (%) – High* Feeding/ beetle-day (%) – Low* Survival at 14 days – High Survival at 14 days – Low LT50 – High (d) LT50 – Low (d) 35.0 (1.5)ab 107.0 (4.9)a 79.0 (3.5)abcd 0.66 (0.03)a 0.47 (0.03)abc 103.0 (3.9)a 73.8 (4.5)abcd 0.3 (0.1)ab 0.7 (0.0)a 10.3 (0.4)b 13.5 (0.1)a 17.2 (1.7)abc 23.1 (0.6)ab 35.8 (3.5)abc 48.0 (1.3)abcd 0.25 (0.02)b 0.34 (0.02)abc 39.4 (3.6)abc 53.2 (3.2)abcd 0.0 (0.0)ab 0.1 (0.0)bc 6.1 (0.2)c 8.7 (0.2)e 5/4 12.8 (1.8)abc 46.3 (7.4)a 26.6 (3.7)abc 134.6 (11.1)a 0.18 (0.02)bc 0.60 (0.08)ab 27.7 (3.9)abc 92.8 (12.8)abc 0.0 (0.0)b 0.3 (0.1)abc 6.1 (0.1)c 11.7 (0.3)b 3c {3,6} 5/5 7.4 (0.6)abc 28.8 (4.5)ab 16.3 (1.4)abc 63.1 (9.9)abcd 0.14 (0.01)cd 0.40 (0.06)abc 21.4 (2.1)abc 62.5 (10.1)abcd 0.0 (0.0)ab 0.4 (0.1)abc 4.6 (0.2)de 10.2 (0.3)bc 3c {n5,6} 5/5 6.6 (0.3)bc 8.8 (1.8)b 14.4 (0.7)bc 19.2 (4.0)cd 0.11 (0.01)cd 0.14 (0.03)c 17.1 (1.9)bc 22.0 (4.3)cd 0.0 (0.0)b 0.1 (0.0)bc 4.3 (0.1)d 6.0 (0.3)fg 3c {n5,n5} 5/5 10.6 (0.7)abc 17.1 (1.7)ab 23.4 (1.6)abc 37.9 (3.8)abcd 0.17 (0.02)bc 0.25 (0.02)abc 27.1 (2.5)abc 39.3 (3.4)abcd 0.0 (0.0)b 0.0 (0.0)c 5.0 (0.2)ef 6.0 (0.2)fg 3c {4,n5} 5/5 6.9 (0.4)bc 16.4 (2.5)ab 15.5 (1.0)bc 37.0 (5.5)bcd 0.09 (0.01)cd 0.22 (0.03)bc 14.4 (0.9)c 34.2 (5.1)bcd 0.2 (0.0)ab 0.2 (0.0)abc 9.4 (0.3)g 9.4 (0.3)cd DEET 5/5 4.7 (0.1)c 7.3 (1.9)b 10.3 (0.2)c 16.1 (4.3)d 0.07 (0.00)d 0.11 (0.03)c 10.7 (0.1)c 17.1 (4.7)d 0.0 (0.0)b 0.2 (0.0)abc 5.2 (0.1)f 7.8 (0.3)def Control 5 48.1 (2.8)a 48.1 (2.8)a 100a 100ab 0.64 (0.03)a 0.64 (0.03)a 100a 100a 0.7 (0.2)a 0.7 (0.2)ab 13.1 (0.5)a 13.1 (0.5)a Methanol 5 40.6 (1.8)ab 40.6 (1.8)a 89.9 (4.1)ab 89.9 (4.1)abc 0.60 (0.02)a 0.60 (0.02)ab 93.7 (3.5)ab 93.7 (3.5)ab 0.4 (0.1)ab 0.4 (0.1)abc 11.6 (0.4)b 11.6 (0.4)b Starvation 1 - - - - - 0.0 6.8 (0.4)c 6.8 (0.4)f - Feeding (%) – High* - - 0.0 *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 73 Figure 4.3. Correlation of the percent feeding by S. zeamais after three days and the average percent survival after 14 days (±SE) for all the compounds tested. There was a significant correlation for both doses tested. 74 Table 4.5. Feeding and mortality (mean ± SE) of T. castaneum in no-choice feeding bioassay on flour disks with different compounds. High dose flour disks were treated with 26 µmol/replicate and low dose flour disks were treated with 8.67 µmol/replicate. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Compound N, H/L Food Food consumed (mg) consumed –High (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High (d) – High – Low LT50 – Low (d) 3c{2,2} 4/5 25.7 (1.2)a 23.8 (0.8)a 113.7 (5.1)a 105.5 (3.4)a 0.34 (0.02)a 0.32 (0.01)a 111.7 (5.6)a 105.0 (3.4)a 1.0 (0.0)abc 1.0 (0.0) 13.9 (0.1)a 14.0 (0.0)a 3c{3,3} 5/5 10.0 (2.4)bc 17.0 (1.0)bc 62.9 (15.1)abc 107.6 (6.4)a 0.13 (0.03)ab 0.23 (0.01)bc 44.1 (10.5)ab 76.1 (4.1)abcd 1.0 (0.0)ab 1.0 (0.0) 14.0 (0.0)a 13.9 (0.1)a 3c{4,4} 5/5 -0.9 (0.7)c 4.8 (1.2)e -5.7 (4.1)c 30.1 (7.8)c -0.01 (0.01)b -3.8 (2.9)b 21.1 (5.4)d 1.0 (0.0)ab 1.0 (0.0) 13.8 (0.2)a 13.8 (0.1)a 3c{3,6} 5/5 4.5 (0.2)bc 21.2 (1.0)ab 20.6 (1.0)bc 97.1 (4.5)a 0.13 (0.01)ab 0.28 (0.01)ab 42.2 (3.3)ab 93.4 (4.6)abc 0.1 (0.1)c 1.0 (0.0) 4.6 (0.3)c --a 3c{n5,6} 5/5 6.6 (0.3)abc 10.9 (1.3)d 30.4 (1.4)abc 50.1 (5.8)bc 0.09 (0.01)ab 0.15 (0.02)d 29.6 (2.3)ab 48.5 (5.7)cd 0.7 (0.0)bc 1.0 (0.0) 11.6 (0.3)b 13.9 (0.0)a 3c{n5,n5} 5/5 6.9 (0.4)abc 12.2 (0.7)cd 36.8 (2.2)abc 65.2 (4.0)b 0.09 (0.01)ab 0.16 (0.01)cd 30.4 (1.8)ab 54.0 (3.3)bcd 0.9 (0.0)abc 1.0 (0.0) 13.8 (0.1)a 13.9 (0.0)a 3c{4,n5} 4/5 6.7 (0.8)abc 12.8 (2.2)cd 29.7 (3.6)abc 56.7 (9.8)bc 0.09 (0.01)ab 0.17 (0.03)cd 29.7 (3.5)ab 57.1 (9.9)abcd 1.0 (0.0)abc 1.0 (0.0) 13.8 (0.2)a 13.8 (0.2)a DEET 5/5 5.8 (0.4)abc 12.8 (0.5)cd 30.8 (1.9)abc 48.4 (5.8)bc 0.08 (0.01)ab 0.17 (0.01)cd 27.1 (2.7)ab 56.8 (2.3)abcd 1.0 (0.0)abc 1.0 (0.0) 13.7 (0.2)a 13.8 (0.2)a Control 5 22.6 (0.8)a 22.6 (0.8)a 100ab 100a 0.30 (0.01)a 0.30 (0.01)a 100ab 0.9 (0.0)abc 0.9 (0.0) 13.7 (0.2)a 13.7 (0.2)a Methanol 5 19.8 (0.7)ab 19.8 (0.7)ab 105.7 (3.6)a 105.7 (3.6)a 0.26 (0.01)a 0.26 (0.01)ab 87.4 (3.0)a 87.4 (3.0)abc 1.0 (0.0)a 1.0 (0.0) --a --a Starvation 1 - - - - - - 0.28 0.28 10.8 (0.5)b 10.8 (0.5)b - 0.06 (0.02)e 100a - *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 75 Figure 4.4. Correlation of the percent feeding by T. castaneum after three days and the average percent survival after 14 days (±SE) for all the compounds tested. There was no significant correlation between feeding and mortality for either dose. 76 Table 4.6. Feeding and mortality (mean ± SE) of T. confusum in no-choice feeding bioassay on flour disks with different compounds. High dose flour disks were treated with 26 µmol/replicate and low dose flour disks were treated with 8.67 µmol/replicate. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Compound N, H/L Food consumed (mg) –High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High (d) – High – Low LT50 – Low (d) 3c{2,2} 4/5 22.2 (1.1)a 26.0 (2.3)a 145.3 (7.3)a 170.6 (14.9)a 0.30 (0.01)a 0.35 (0.03)a 145.9 (12.8)a 145.9 (6.4)a 1.0 (0.0) 1.0 (0.0) --b --b 3c{3,3} 5/5 15.9 (1.2)ab 24.7 (1.3)a 92.9 (7.1)ab 144.1 (7.8)a 0.21 (0.02)ab 0.34 (0.01)a 103.9 (17.6)ab 103.9 (7.9)ab 1.0 (0.0) 1.0 (0.0) 13.9 (0.1)b --b 3c{4,4} 5/5 4.5 (0.4)abc 6.3 (1.4)b 26.5 (2.5)abc 36.7 (8.1)c 0.06 (0.01)ab 0.08 (0.02)b 29.6 (6.2)abc 29.6 (2.8)abc 1.0 (0.0) 1.0 (0.0) 13.9 (0.1)b 14.0 (0.0)b 3c{3,6} 5/5 2.7 (0.2)bc 15.6 (1.2)ab 16.1 (0.5)bc 94.3 (7.1)b 0.04 (0.00)b 0.21 (0.02)ab 17.3 (1.2)bc 17.3 (0.5)bc 1.0 (0.0) 1.0 (0.0) 13.8 (0.2)b --b 3c{n5,6} 5/5 5.0 (0.8)abc 4.4 (1.3)b 30.2 (5.0)abc 26.4 (7.7)c 0.07 (0.01)ab 0.06 (0.02)b 32.6 (12.2)abc 32.6 (5.5)abc 1.0 (0.0) 1.0 (0.0) 14.0 (0.0)b 14.0 (0.0)b 3c{n5,n5} 5/5 3.8 (1.3)abc 11.7 (0.7)ab 21.6 (7.3)abc 66.8 (4.0)bc 0.05 (0.02)ab 0.16 (0.01)ab 25.4 (19.4)abc 25.5 (8.7)abc 0.9 (0.0) 1.0 (0.0) 13.8 (0.1)b --b 3c{4,n5} 5/5 4.4 (1.2)abc 9.7 (1.5)ab 28.6 (7.6)abc 63.7 (10.0)bc 0.06 (0.02)ab 0.13 (0.02)ab 28.9 (16.5)abc 28.9 (7.4)abc 1.0 (0.0) 1.0 (0.0) 13.9 (0.1)b 14.0 (--)b DEET 5/5 2.3 (0.7)c 8.6 (0.8)b 13.3 (4.1)c 48.8 (4.8)c 0.03 (0.01)b 15.3 (10.5)c 15.3 (4.7)c 1.0 (0.0) 1.0 (0.0) --b 14.0 (0.0)b Control 5 15.3 (3.0)abc 15.3 (3.0)ab 100a 100b 0.21 (0.04)ab 0.21 (0.04)ab 100ab 100ab 1.0 (0.0) 1.0 (0.0) 13.8 (0.2)b 13.8 (0.2)b Methanol 5 11.9 (2.7)abc 11.9 (2.7)ab 67.6 (15.3)abc 67.6 (15.3)bc 0.16 (0.04)ab 0.16 (0.04)ab 77.3 (39.0)abc 77.3 (17.4)abc 1.0 (0.0) 1.0 (0.0) --b --b Starvation 1 - - - - - 0.44 0.44 13.0 (0.3)a 13.0 (0.3)a - 0.11 (0.01)b - - *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 77 Figure 4.5. Correlation of the percent feeding by T. confusum after three days and the average percent survival after 14 days (±SE) for all the compounds tested. There was no significant correlation between feeding and mortality for either dose. 78 Table 4.7. Feeding and mortality (mean ± SE) of R. dominica in no-choice feeding bioassay on flour disks with different compounds. High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Compound N, H/L Food consumed (mg) –High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High 3c{2,2} 4/4 0.1 (0.0)cd 5.3 (0.7)abc 0.1 (0.0)bc 86.3 (11.9)abc 3c{3,3} 5/5 0.6 (0.3)bcd 6.0 (0.5)abc 7.5 (4.0)bc 3c{4,4} 5/5 1.1 (0.1)abcd -0.1 (0.1)c 3c{3,6} 5/5 2.1 (0.1)abcd 3c{n5,6} 4/5 3c{n5,n5} Food consumed (mg/beetleday) – Low Feeding/ beetle-day (%) – High* Feeding/ beetle-day (%) – Low* 0.00 (0.00)cd 0.07 (0.01)abc 1.3 (0.1)cd 70.3 (6.3)abcd 0.01 (0.00)bcd 0.08 (0.01)abc 12.4 (1.4)abc -1.4 (1.0)cd 0.02 (0.00)abcd 3.7 (1.0)abc 26.7 (1.8)abc 2.6 (0.6)abcd 0.9 (0.1)abc 4/5 1.9 (0.2)abcd 3c{4,n5} 4/5 DEET Survival at 14 days – High Survival at 14 days – Low LT50 – High (d) LT50 – Low (d) 66.0 (8.7)ab 0.9 (0.1)a 0.9 (0.0)a 13.2 (0.2)a 13.6 (0.2)a 7.8 (4.3)bcd 71.4 (6.1)ab 0.2 (0.1)abc 0.8 (0.0)a 7.2 (0.3)c 13.6 (0.2)a -0.00 (0.00)c 16.7 (1.9)abcd -1.5 (1.1)b 0.0 (0.0)abc 0.1 (0.0)c 5.1 (0.2)e 6.8 (0.3)ce 48.3 0.04 (13.4)abcd (0.01)abcd 0.05 (0.01)abc 38.3 (4.9)abcd 44.8 (12.1)ab 0.0 (0.0)abc 0.8 (0.0)ab 4.7 (0.2)ef 12.9 (0.2)ab 33.5 (7.4)abc 12.2 (1.1)abcd 0.05 (0.01)abcd 0.02 (0.00)abc 47.4 (11.6)abcd 15.5 (1.3)ab 0.0 (0.0)c 0.1 (0.0)c 4.4 (0.1)f 5.3 (0.3)d 0.7 (0.1)bc 24.5 (2.1)abc 8.8 (1.7)bcd 0.03 (0.00)abcd 0.01 (0.00)bc 27.0 (3.0)abcd 9.0 (2.0)b 0.0 (0.0)abc 0.1 (0.0)c 6.1 (0.3)d 7.2 (0.3)c -1.7 (0.3)d -0.2 (0.1)c -28.6 (5.1)c -3.0 (1.0)c -0.04 (0.01)d -0.00 (0.00)c -33.6 (6.1)d -3.5 (1.3)b 0.0 (0.0)bc 0.1 (0.0)c 4.6 (0.2)ef 5.3 (0.3)de 4/5 8.3 (0.5)abc 4.5 (0.9)abc 107.0 (6.3)a 57.5 0.11 (11.9)abcd (0.01)abc 0.06 (0.01)abc 98.7 (5.7)ab 52.6 (10.8)ab 0.7 (0.1)abc 0.7 (0.1)b 12.9 (0.2)b 12.9 (0.2)b Control 5 8.6 (0.9)a 8.6 (0.9)ab 100ab 100ab 0.11 (0.01)ab 0.11 (0.01)ab 100abc 100c 0.9 (0.0)a 0.9 (0.0)a 13.8 (0.1)a 13.8 (0.1)a Methanol 5 8.8 (0.4)ab 8.8 (0.4)a 113.7 (5.1)a 113.7 (5.1)a 0.12 (0.01)a 0.12 (0.01)a 107.7 (6.0)a 107.7 (6.0)a 0.8 (0.0)ab 0.8 (0.0)a 13.8 (0.1)a 13.8 (0.1)a Starvation 1 - - - - - - - - 0.0 0.0 9.6 (0.4)c 9.6 (0.4)e *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 79 Figure 4.6. Correlation of the percent feeding by R. dominica after three days and the average percent survival after 14 days (±SE) for all the compounds tested. There was a significant relationship between feeding and mortality for both dose. 80 3c{2,2}which was not significantly different from the controls (High: χ210 = 1112.37, P < 0.001; Low: χ210 = 1142.00, P < 0.001). Several compounds that did not have an effect on feeding did have significantly different LT50 than the controls (Table 4.7). There was also a significant correlation between amount of flour disk eaten and mortality for both doses (High: r = 0.62, P < 0.001, n = 48; Low: r = 0.86, P < 0.001, n = 49). (Figure 4.6). There were significant differences (P < 0.05) among species in mean amount consumed, mortality, and the LT50 (Tables 4.8-4.17). Because the difference between the mean amounts of disk consumed by each species could be related to the significant differences in species weight, the amount consumed was standardized to each species’ weight (Tables 4.18, 4.19). At the tested high dose, R. dominica consumed the least amount of flour disk (% feeding per beetle day) for treatment 3c{2,2} (F4,15 = 1688.23, P < 0.001) and 3c{3,3} (F4,19 = 28.19, P < 0.001) when compared to the other four species (Tables 4.10, 4.11, 4.18). Sitophilus oryzae and S. zeamais often consumed the same amount of flour disks except more compounds reduced feeding by S. zeamais (Tables 4.3, 4.4). Tribolium castaneum and T. confusum usually consumed the same amount of flour disks except for 3c{4,4} (H4 = 20.66, P < 0.001) (Table 4.12) and 3c{3,6} (F4,20 = 36.08, P < 0.001) (Table 4.13). At the low dose, there was more overlap between the amounts of flour disk eaten by the five species (Table 4.19). However, R. dominica and S. zeamais consumed the least amount of flour disks treated with 3c{3,3} (F4,19 = 29.62, P < 0.001) (Table 4.11). Rhyzopertha dominica also consumed the least amount of flour disks treated with 3c{n5,n5} (F4,20 = 103.20, P < 0.001) (Table 4.15). Sitophilus zeamais was the least deterred from eating by 3c{4,4} (F4,19 = 29.62, P < 0.001) (Table 4.12) while T. castaneum ate the most of the flour disks treated with 3c{n5,6} (F4,20 = 11.04, P < 0.001) (Table 4.14). 81 Survival was also significantly different (P < 0.05) between the closely related species although not as noticeable as the difference in feeding deterrence (Tables 4.8-4.17). Sitophilus zeamais had a significantly higher LT50 than S. oryzae on disks treated with 3c{4,4} (High: χ24 = 702.54, P < 0.001; Low: χ24 = 717.56, P < 0.001) and 3c{4,n5} (High: χ24 = 659.98, P < 0.001; Low: χ24 = 600.50, P < 0.001) (Tables 4.12, 4.16). Tribolium castaneum had a significantly lower LT50 on the high dose of 3c{3,6} when compared with T. confusum (High: χ24 = 379.40, P < 0.001; Low: χ24 = 211.33, P < 0.001) (Table 4.13). There were significant differences between the species in both the control and the methanol control feeding bioassays (Control: χ24 = 118.20, P < 0.001; Methanol Control: χ24 = 210.44, P < 0.001). The Sitophilus spp. had the lowest LT50 in both controls (Table 4.8, 4.9) and for several of the treatments, Sitophilus spp. had the lowest LT50 for many of the compounds [3c{2,2} (High: χ24 = 209.63, P < 0.001; Low: χ24 = 155.09, P < 0.001), 3c{3,3} (High: χ24 = 520.09, P < 0.001; Low: χ24 = 571.01, P < 0.001), 3c{n5,6} (High: χ24 = 569.21, P < 0.001; Low: χ24 = 581.03, P < 0.001), 3c{n5,n5} (High: χ24 = 588.21, P < 0.001; Low: χ24 = 60.51, P < 0.001), and DEET (High: χ24 = 820.82, P < 0.001; Low: χ24 = 600.50, P < 0.001)] but not all (e.g. 3c{4,4}). Rhyzopertha dominica had a lower LT50 than S. zeamais for the disks treated with 3c{4,4} (Table 4.12) and lower than S. oryzae for disks treated with 3c{4,n5} (Table 4.16). Tribolium confusum was the only species tested at the high dose of 3c{3,6} that had an LT50 higher than five days (Table 4.13). There were no significant correlations (P > 0.05) for any of the five species tested between the Log Kow and the percent feeding reduction per beetle-day. There were also no signficant correlations (P > 0.05) between the Log Kow and the mortality for the high dose in any of the five species tested. 82 Table 4.8. Summary table for control disk no-choice feeding bioassay (mean ± SE). Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N Food consumed (mg) % Feeding* Mean amount consumed (mg/beetle-day) Feeding/ beetle-day (%)* Survival at 14 days LT50 (d) S. oryzae 5 31.0 (0.8)ab 100 0.42 (0.01)ab 100 0.5 (0.2)b 13.4 (0.1)c S. zeamais 5 48.1 (2.8)a 100 0.64 (0.03)a 100 0.7 (0.2)ab 13.1 (0.2)c T. castaneum 5 22.6 (0.8)abc 100 0.30 (0.01)abc 100 0.9 (0.0)ab 13.7 (0.2)ab T. confusum 5 15.3 (3.0)bc 100 0.21 (0.04)bc 100 1.0 (0.0)a 13.8 (0.2)b R. dominica 5 8.6 (0.9)c 100 0.11 (0.01)c 100 0.9 (0.0)ab 13.8 (0.1)a *Percent feeding on the control disks was set at 100%. No further analyses were performed on these data. 83 Table 4.9. Summary table for methanol control disk no-choice feeding bioassay (mean ± SE). The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N Food consumed (mg) % Feeding* Mean amount consumed (mg/beetle-day) Feeding/ beetle-day (%)* Survival at 14 days LT50 (d) S. oryzae 5 27.8 (1.4)ab 93.8 (4.7)ab 0.38 (0.01)ab 91.7 (2.6) 0.7 (0.2)ab 13.3 (0.2)c S. zeamais 5 40.6 (1.8)a 89.9 (4.1)ab 0.60 (0.02)a 93.7 (3.5) 0.4 (0.1)b 11.6 (0.4)d T. castaneum 5 19.8 (0.7)abc 105.7 (3.6)a 0.26 (0.01)abc 87.4 (3.0) 1.0 (0.0)a --a T. confusum 5 11.9 (2.7)bc 67.6 (15.3)b 0.16 (0.04)bc 77.3 (17.4) 1.0 (0.0)a --a R. dominica 5 8.8 (0.4)c 113.7 (5.1)a 0.12 (0.01)c 107.7 (6.0) 0.8 (0.0)ab 13.8 (0.1)b *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 84 Table 4.10. Summary table for test compound 3c{2,2} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N Food (H/L) consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low S. oryzae 4/5 31.6 (1.6)b 9.0 (3.0)a 101.8 (5.1)b 107.3 (11.7)b 0.42 (0.02)a S. zeamais 4/5 47.4 (2.2)a 17.2 (3.8)a 107.0 (4.9)b 79.0 (7.8)b 25.7 (1.2)bc 10.0 (5.3)b 113.7 (5.1)b T. castaneum 4/5 Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High – High – Low (d) LT50 – Low (d) 0.45 (0.06)ab 100.0 (3.8)b 108.0 (15.4)b 0.6 (0.2)ab 0.5 (0.5)ab 12.4 (0.3)c 13.1 (0.2)d 0.66 (0.03)a 0.47 (0.06)a 103.0 (3.9)b 73.8 (10.0)c 0.3 (0.1)b 0.7 (0.1)b 10.3 (0.4)d 13.5 (0.1)c 105.5 (7.7)b 0.34 (0.02)ab 0.32 (0.02)c 111.7 (5.6)b 105.0 (7.7)b 1.0 (0.0)a 1.0 (0.0)ab 13.9 (0.1)a 14.0 (0.0)ab 1.0 (0.0)a --a T. confusum 4/5 22.2 (1.1)c 15.9 (2.7)b 145.3 (7.3)a 170.6 (33.3)a 0.30 (0.01)ab 0.35 (0.07)bc 145.9 (6.4)a 171.1 (32.7)a 1.0 (0.0)a R. dominica 4/4 0.1 (0.0)d 0.6 (0.8)c 0.1 (0.0)c 86.3 (23.8)b 0.00 (0.00)b 0.07 (0.02)d 66.0 (17.5)c 0.9 (0.1)ab 0.9 (0.0)ab 13.2 (0.2)b 1.3 (0.1)c --a *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 85 13.6 (0.2)b Table 4.11. Summary table for test compound 3c{3,3} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N (H/L) Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High – High – Low (d) LT50 – Low (d) S. oryzae 4/5 9.0 (3.0)b 24.3 (2.2)a 30.3 (10.0)bc 81.9 (7.5)c 0.12 (0.04)b 0.33 (0.04)a 29.4 (8.6)bc 78.8 (9.2)b 0.0 (0.0)b 0.7 (0.1)c 5.8 (0.2)c 12.9 (0.2)c S. zeamais 5/5 17.2 (3.8)a 23.1 (1.4)a 35.8 (7.9)b 48.0 (2.8)d 0.25 (0.05)a 0.34 (0.05)a 39.4 (8.1)b 53.2 (7.1)c 0.0 (0.1)b 0.1 (0.0)d 6.1 (0.2)c 8.7 (0.2)d T. castaneum 5/5 10.0 (5.3)b 17.0 (2.3)b 62.9 (33.8)ab 107.6 (14.4)b 0.13 (0.07)b 0.23 (0.03)b 44.1 (23.4)b 76.1 (9.2)b 1.0 (0.0)a 1.0 (0.0)ab 14.0 (0.0)a 13.9 (0.1)a T. confusum 5/5 15.9 (2.7)ab 24.7 (3.0)a 92.9 (15.8)a 144.1 (17.5)a 0.21 (0.04)ab 0.34 (0.02)a 103.9 (17.6)a 167.7 (10.3)a 1.0 (0.0)a 1.0 (0.0)a 13.9 (0.1)a --a R. dominica 5/5 0.6 (0.8)c 7.5 (9.0)c 70.3 (14.1)cd 0.01 (0.01)c 7.8 (9.5)c 71.5 (13.6)bc 0.2 (0.1)ab 0.8 (0.0)b 7.2 (0.3)b 13.6 (0.1)b 6.0 (1.2)c 0.08 (0.02)c *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 86 Table 4.12. Summary table for test compound 3c{4,4} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N (H/L) Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High S. oryzae 5/5 0.7 (2.1)b 2.3 (1.6)bc 2.3 (7.1)ab 7.7 (5.4)cd S. zeamais 5/5 12.8 (4.0)a 46.3 (14.7)a 26.6 (8.3)ab T. castaneum 5/5 -0.9 (1.5)b 4.8 (2.7)bc T. confusum 5/5 4.5 (1.0)ab R. dominica 5/5 1.1 (0.3)ab Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High – High – Low (d) LT50 – Low (d) 0.01 (0.03)ab 0.03 (0.02)ab 2.3 (7.6)bc 8.3 (5.5)cd 0.0 (0.0)b 0.0 (0.0)c 4.2 (0.1)d 4.4 (0.1)d 134.6 (22.2)a 0.18 (0.06)a 0.60 (0.16)a 27.7 (8.6)ab 92.8 (25.7)a 0.0 (0.0)b 0.3 (0.1)abc 6.1 (0.1)b 11.7 (0.3)b -5.7 (9.2)b 30.1 (17.3)bc -0.01 (0.02)b 0.06 (0.04)ab -3.8 (6.5)c 21.1 (12.1)bc 1.0 (0.0)a 1.0 (0.0)ab 13.8 (0.2)a 13.8 (0.1)a 6.3 (3.1)b 26.5 (5.6)a 36.7 (18.0)b 0.06 (0.01)ab 0.08 (0.04)ab 29.6 (6.2)a 41.1 (20.2)b 1.0 (0.1)a 1.0 (0.0)a 13.9 (0.1)a 14.0 (0.0)a -0.1 (0.2)c 12.4 (3.2)ab -1.4 (2.2)d 0.02 (0.0)ab -1.5 (2.4)d 0.0 (0.1)ab 0.1 (0.1)bc 5.1 (0.2)c 6.8 (0.3)c -0.00 (0.00)b Feeding/ beetleday (%) – High* 16.7 (1.9)abc *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 87 Table 4.13. Summary table for test compound 3c{3,6} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N (H/L) Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High Feeding/ beetleday (%) – Low Survival Survival LT50 – at 14 days at 14 days High – High* – Low* (d) LT50 – Low (d) S. oryzae 5/5 3.0 (0.3)c 11.9 (6.2)c 9.8 (0.8)c 38.2 (19.9)b 0.07 (0.01)b 0.17 (0.09)b 15.8 (1.5)b 39.7 (20.6)c 0.0 (0.0)b 0.7 (0.1)ab 3.7 (0.1)c 11.5 (0.4)c S. zeamais 5/5 7.4 (1.4)a 28.8 (10.1)a 16.3 (3.1)b 63.1 (22.1)ab 0.14 (0.03)a 0.40 (0.14)a 21.4 (4.6)b 62.5 (22.5)bc 0.0 (0.0)b 0.4 (0.2)b 4.6 (0.2)b 10.2 (0.3)d T. castaneum 5/5 4.5 (0.5)b 21.2 (2.2)ab 20.6 (2.2)b 97.1 (10.1)a 0.13 (0.02)a 0.28 (0.03)ab 42.2 (7.3)a 93.4 (10.3)ab 0.1 (0.1)ab 1.0 (0.0)a 4.6 (0.3)bc --a T. confusum 5/5 2.7 (0.2)cd 15.6 (2.6)bc 16.1 (1.1)b 94.3 (15.9)a 0.04 (0.00)b 0.21 (0.04)b 17.3 (1.2)b 101.8 (17.1)a 1.0 (0.0)a 13.8 (0.2)a 14.0 (0.0)a R. dominica 5/5 2.1 (0.3)d 3.7 (2.3)d 48.3 (30.0)b 0.04 (0.01)b 0.05 (0.03)c 38.3 (11.0)a 44.8 (27.0)c 0.0 (0.1)ab 0.8 (0.1)ab 4.7 (0.2)b 12.9 (0.2)b 26.8 (4.0)a 1.0 (0.0)a *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 88 Table 4.14. Summary table for test compound 3c{n5,6} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N (H/L) Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High – High – Low (d) LT50 – Low (d) S. oryzae 5/5 3.9 (0.6)bc 3.6 (1.6)c 12.7 (2.1) 11.7 (5.2)b 0.07 (0.01)b 0.06 (0.03)b 15.8 (2.7) 13.7 (6.0)b 0.0 (0.0)b 0.0 (0.0)b 4.2 (0.1)d 4.7 (0.1)c S. zeamais 5/5 6.6 (0.7)a 8.8 (4.1)ab 14.4 (1.6) 19.2 (8.9)b 0.11 (0.03)a 0.14 (0.06)a 17.1 (4.2) 22.0 (9.5)b 0.0 (0.0)b 0.1 (0.0)ab 4.3 (0.1)cd 6.0 (0.3)b T. castaneum 5/5 6.6 (0.7)a 10.9 (2.8)a 30.4 (3.2) 50.1 (12.9)a 0.09 (0.02)ab 0.15 (0.04)a 29.6 (5.2) 48.5 (12.8)a 0.7 (0.1)ab 1.0 (0.0)a 11.6 (0.3)b 13.9 (0.0)a T. confusum 5/5 5.0 (1.9)ab 4.4 (2.8)bc 30.2 (11.3) 26.4 (17.2)b 0.07 (0.02)ab 0.06 (0.04)b 32.6 (12.2) 28.6 (18.6)ab 1.0 (0.0)a 1.0 (0.0)a 14.0 (0.0)a --a R. dominica 4/5 2.6 (1.3)c 0.9 (0.2)c 33.5 (16.5) 12.2 (2.4)b 0.05 (0.03)b 47.4 (26.0) 15.5 (2.8)b 0.0 (0.0)b 0.1 (0.1)ab 4.4 (0.1)c 0.02 (0.00)b *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 89 5.3 (0.3)b Table 4.15. Summary table for test compound 3c{n5,n5} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N (H/L) Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High S. oryzae 4/5 5.3 (0.6)ab 6.7 (1.4)c 17.7 (2.0)b 22.8 (4.7)c S. zeamais 5/5 10.6 (1.6)a 17.1 (3.8)a 23.4 (3.6)ab T. castaneum 5/5 6.9 (0.9)ab 12.2 (1.7)b T. confusum 5/5 3.8 (2.9)b R. dominica 4/5 1.9 (0.3)b Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival Survival at LT50 – at 14 days 14 days – High – High Low (d) LT50 – Low (d) 0.08 (0.01)ab 0.09 (0.02)c 19.1 (2.3) 22.2 (4.3)d 0.0 (0.0)b 0.0 (0.0)b 4.9 (0.1)c 5.8 (0.1)c 37.9 (8.5)b 0.17 (0.02)a 0.25 (0.05)a 27.1 (5.5) 39.3 (7.5)c 0.0 (0.0)b 0.0 (0.0)b 5.0 (0.2)c 6.0 (0.2)c 36.8 (4.8) a 65.2 (8.9)a 0.09 (0.01)ab 0.16 (0.02)b 30.4 (4.0) 54.0 (7.4)b 0.9 (0.0)a 13.8 (0.1)a 13.9 (0.0)a 11.7 (1.6)b 21.6 (16.4)ab 66.8 (8.9)a 0.05 (0.02)b 0.16 (0.01)b 25.5 (19.4) 78.6 (5.5)a 0.9 (0.0)ab 1.0 (0.0)a 13.8 (0.1)a --a 0.7 (0.3)d 24.5 (4.2)ab 8.8 (3.9)d 0.03 (0.00)b 0.01 (0.01)d 27.0 (6.0) 9.0 (4.5)e 0.9 (0.0)ab 0.08 (0.1)ab 6.1 (0.3)b 7.2 (0.3)b 1.0 (0.0)a *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 90 Table 4.16. Summary table for test compound 3c{4,n5} no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. Statistically significant differences in columns (P < 0.05) between treatments are represented by different letters. Species N, H/L Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/beetleday) – High Food consumed (mg/beetleday) – Low S. oryzae 5/4 -0.9 (0.6)ab 0.3 (0.5)b -3.0 (108)ab 1.0 (1.5)ab -0.02 (0.01)ab S. zeamais 5/5 6.9 (1.0)a 16.4 (5.5)a 15.5 (1.0)ab 37.0 (12.4)ab T. castaneum 4/5 6.7 (1.6)a 12.8 (4.9)a 29.7 (7.3)a T. confusum 5/5 4.4 (2.6)ab 9.7 (3.4)a R. dominica 5/5 -1.7 (0.7)b -0.2 (0.1)b Feeding/ beetleday (%) – Low* Survival Survival LT50 – at 14 days at 14 days High – High – Low (d) LT50 – Low (d) 0.00 (0.01)ab -3.8 (2.3)ab 1.2 (1.6)ab 0.0 (0.0)b 0.0 (0.0)c 4.7 (0.2)c 5.1 (0.2)c 0.09 (0.01)a 0.22 (0.07)a 14.4 (2.0)ab 34.2 (11.5)ab 0.2 (0.1)ab 0.2 (0.1)b 9.4 (0.3)b 9.4 (0.3)b 56.7 (21.9)a 0.09 (0.02)a 0.17 (0.07)a 29.7 (7.0)a 57.1 (22.2)a 1.0 (0.1)a 1.0 (0.0)a 13.8 (0.2)a 13.8 (0.2)a 28.6 (7.6)a 63.7 (22.5)a 0.06 (0.03)ab 0.13 (0.05)ab 28.9 (16.5)a 63.4 (22.3)a 1.0 (0.0)a 1.0 (0.0)a 13.9 (0.1)a 14.0 (-)a -28.6 (5.1)b -3.0 (2.1)b -0.04 (0.02)b -3.5 (2.9)b 0.0 (0.0)b 0.1 (0.1)c 4.6 (0.2)c 5.3 (0.3)c -0.00 (0.00)b Feeding/ beetleday (%) – High* -33.6 (13.7)b *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 91 Table 4.17. Summary table for test compound DEET no-choice feeding bioassay (mean ± SE). High dose flour disks were treated with 26 µmol per replicate and low dose flour disks had 1/3 the concentration of the high dose. The median lethal time (LT50) could not be calculated for all treatments as there was not enough mortality. Statistically significant differences in columns (P < 0.05) between treatments were represented by different letters. Species N, H/L Food consumed (mg) – High Food consumed (mg) – Low Feeding (%) – High* Feeding (%) – Low* Food consumed (mg/ beetleday) – High Food consumed (mg/beetleday) – Low Feeding/ beetleday (%) – High* Feeding/ beetleday (%) – Low* Survival at 14 days – High Survival LT50 – at 14 days High – Low (d) LT50 – Low (d) S. oryzae 5/5 2.7 (0.3)b 8.4 (8.7)ab 9.2 (1.0)b 28.3 (29.4)b 0.04 (0.00)b 0.11 (0.11)ab 9.8 (0.8)b 0.4 (0.4)b 0.0 (0.0)b 0.2 (0.2)c 4.2 (0.1)e 7.2 (0.3)c S. zeamais 5/5 4.7 (0.1)ab 7.3 (4.3)ab 10.3 (0.5)b 16.1 (9.6)c 0.07 (0.00)ab 0.11 (0.07)ab 10.7 (0.2)ab 17.1 (10.4)ab 0.0 (0.0)b 0.2 (0.1)bc 5.2 (0.1)d 7.8 (0.3)c T. castaneum 5/5 5.8 (0.8)ab 12.8 (1.1)a 30.8 (4.2)ab 48.4 (13.1)a 0.08 (0.02)ab 0.17 (0.02)a 27.1 (5.9)ab 56.8 (5.1)a 1.0 (0.0)a 1.0 (0.0)ab 13.7 (0.2)b 13.8 (0.2)a T. confusum 5/5 2.3 (0.7)b 8.6 (1.9)ab 13.3 (9.2)ab 48.8 (10.7)a 0.03 (0.02)b 0.11 (0.03)ab 15.3 (10.5)b 55.4 (12.3)a 1.0 (0.0)a 1.0 (0.0)a --a 14.0 (0.0)a R. dominica 5/5 8.3 (0.5)a 4.5 (2.1)b 107.0 (14.0)a 57.5 (16.5)a 0.11 (0.01)a 0.06 (0.03)b 52.6 (24.2)a 0.7 (0.1)ab 0.7 (0.1)abc 12.9 (0.2)c 12.9 (0.2)b 98.7 (12.8)a *Feeding on control plates was set at 100%. The same control plates were used for both the high and low dose analyses. 92 Table 4.18. Mean amount consumed (mg/beetle-day) divided by the mean insect weight in the no-choice feeding bioassay at the high treatment dose. Statistically significant differences in columns (P < 0.05) are represented by different letters. Species 3c{2,2} 3c{3,3} 3c{4,4} 3c{3,6} 3c{n5,6} 3c{n5,n5} 3c{4,n5} DEET Control Methanol S. oryzae 0.26 (0.01)a 0.08 (0.02)a 0.01 (0.02)ab 0.04 (0.00)b 0.04 (0.01) 0.05 (0.01)a -0.01 (0.01)bc 0.03 (0.00)abc 0.26 (0.02)a 0.24 (0.02)a S. zeamais 0.14 (0.01)b 0.05 (0.01)a 0.04 (0.01)a 0.03 (0.01)b 0.02 (0.01) 0.04 (0.01)ab 0.02 (0.00)ab 0.01 (0.00)bc 0.13 (0.01)b 0.13 (0.01)ab T. castaneum 0.16 (0.01)b 0.06 (0.03)a -0.01 (0.01)b 0.06 (0.01)a 0.04 (0.01) 0.04 (0.01)ab 0.04 (0.01)a 0.04 (0.01)ab 0.14 (0.01)b 0.13 (0.01)ab T. confusum 0.11 (0.00)c 0.08 (0.01)a 0.02 (0.00)ab 0.01 (0.00)c 0.02 (0.01) 0.02 (0.01)b 0.02 (0.01)ab 0.01 (0.01)c 0.07 (0.03)c 0.06 (0.03)b R. dominica 0.00 (0.00)d 0.01 (0.01)b 0.01 (0.00)ab 0.03 (0.01)b 0.04 (0.02) 0.02 (0.01)b -0.03 (0.01)c 0.09 (0.01)a 0.09 (0.02)c 0.09 (0.01)b 93 Table 4.19. Mean amount consumed (mg/beetle-day) divided by the mean insect weight in the no-choice feeding bioassay at the low treatment dose (1/3 the concentration of the high dose). Statistically significant differences in columns (P < 0.05) are represented by different letters. Species 3c{2,2} 3c{3,3} 3c{4,4} 3c{3,6} 3c{n5,6} 3c{n5,n5} 3c{4,n5} DEET Control Methanol S. oryzae 0.28 (0.04)a 0.21 (0.02)a 0.02 (0.01)b 0.10 (0.05)ab 0.04 (0.02)b 0.06 (0.01)b 0.00 (0.00)b 0.07 (0.07)ab 0.26 (0.02)a 0.24 (0.02)a S. zeamais 0.10 (0.01)cd 0.07 (0.01)c 0.12 (0.03)a 0.08 (0.03)ab 0.03 (0.01)b 0.05 (0.01)b 0.05 (0.02)ab 0.02 (0.01)b 0.13 (0.01)b 0.13 (0.01)ab T. castaneum 0.15 (0.01)b 0.11 (0.01)b 0.03 (0.02)b 0.13 (0.01)a 0.07 (0.02)a 0.08 (0.01)a 0.08 (0.03)a 0.08 (0.01)a 0.14 (0.01)b 0.13 (0.01)ab T. confusum 0.13 (0.01)bc 0.12 (0.01)b 0.03 (0.01)b 0.07 (0.01)ab 0.02 (0.01)b 0.06 (0.00)b 0.05 (0.02)ab 0.04 (0.01)ab 0.07 (0.03)c 0.06 (0.03)b R. dominica 0.06 (0.01)d 0.06 (0.01)c -0.00 (0.00)b 0.04 (0.02)b 0.01 (0.00)b 0.01 (0.00)c -0.00 (0.00)b 0.05 (0.02)ab 0.09 (0.02)c 0.09 (0.01)b 94 Discussion Closely related species There were significantly different responses between the two sets of closely related species. For example, at the low dose, DEET reduced feeding in both species of Sitophilus, while S. oryzae feeding was reduced by 3c{4,n5} and S. zeamais feeding was reduced by 3c{n5,6}. Standardized for the weight of the insect, high dose 3c{3,6} reduced the feeding for T. confusum significantly more compared with the feeding by T. castaneum. Antifeedants, in general, have more differential bioactivity than insecticides (Isman, 2002), and variation in the effects of antifeedants for closely related species also has been observed. For example, Trichoplusia ni was more sensitive to botanical antifeedants than Pseudaletia unipuncta, even though both are noctuid caterpillars (Akhtar et al., 2008), and two species of Callosobruchus leaf beetles showed variation in tolerance to essential oils (Nyamador et al., 2009). While my prediction that closely related species would show similar responses is supported by some of the feeding and mortality results, they were not consistent. There are clearly enough differences within these groups to result in detectable variation in their responses to the same compounds. In many bioassays designed to evaluate feeding deterrence and toxicity of a wide range of compounds including essential oils, one species in each of several genera are tested, which allows for a wide range of species to be assessed in an efficient manner (Hou and Fields, 2003; Xie et al., 1995; Huang et al., 1997; Shaaya et al., 1997; Suthisut et al., 2011). However, based on the variation observed in this feeding bioassay, some caution should be taken when extrapolating the effect of a compound, even to extremely closely related species. This was 95 observed not only in feeding deterence but in mortality effects. Assuming otherwise could result in ineffective control strategies even if both related species are sympatric. Primary vs. secondary pests My bioassays show that the primary pests (Sitophilus spp., R. dominica) exhibited a greater response to the test compounds (both in regards to reduced feeding and increased mortality) than the secondary pests (T. castaneum and T. confusum), supporting my prediction as primary pests in general have a narrower host range than secondary pests (Rees, 2004) and thus may not be adapted to handle a wide range of secondary metabolites. Tribolium canstaneum and T. confusum have been found in a wide range of materials: grain, flour, processed cereal products, beans, dried fruits and vegetables, chocolate, spices, and even museum specimens (Mason, 2003). There is evidence that cannibalism allows T. castaneum to colonize new environments that are not nutritionally rich (Via, 1999) and Tribolium can develop and reproduce in environments that are not suitable for other insect pests (Hall, 1970). Rhyzopertha dominica has been found in grains, beans, books, and packing material made from wood, although it is most reproductively successful on wheat (Eddie, 2012). Sitophilus oryzae and S. zeamais are found in whole kernels and seeds, rarely in milled products and nuts and dried fruits (Hahn et al., 2016). The primary pests need whole grains to reproduce and therefore have evolved to specialize in finding and using these whole grain resources. In comparison, the secondary pests need to find damaged grains as they are incapable of using whole grains. Sitophilus orzyae showed attraction to volatiles that are characteristic of fresh grain while T. castaneum responded to volatiles from the grain that may signal older and damaged grain substrates (Phillips et al., 1993). I therefore propose that the requirement of whole grains for 96 these primary species to reproduce has resulted in them being more specialized in their hosts, having to find pockets of whole grains. On the other hand, I propose that secondary pests being able to use damaged grains and processed resources both to feed and reproduce in has resulted in these insects finding and being exposed to a wide range of food sources (more generalist) and coming into contact with a more diverse range of chemicals in more wide-ranging their foraging and feeding habits. The genome of T. castaneum lends support to this interpretation (Tribolium Genome Sequencing Consortium, 2008). Aspects of the full genome indicate that the insect has a large number of genes for odorant and gustatory receptors. The genome also reveals a large number of genes for cytochrome P450s which are often associated with detoxification of exogenous compounds causing speculation that Tribolium spp. have adapted to diverse chemical environments through this expansion of the cytochrome P450 groups (Tribolium Genome Sequencing Consortium, 2008). The apparent ability to detect a wide range of environmental chemical stimuli and to deal with a similar range of toxins could be because the species utilizes a wide range of food sources. Generalist herbivores tend to possess the ability to detoxify a wide range of plant defensive chemicals and can sometimes better handle compounds not previously encountered compared to specialist herbivores (Castells and Berenbaum, 2008; Haylon et al., 2015). This ability by Tribolium spp. to handle novel chemistry compared to Sitophilus spp. and, to a lesser extent, R. dominica was supported by my bioassays. There are significant differences in the toxicity of various pesticide treatments to different species (Huang and Ho, 1998; Athanassiou et al., 2005; Chu et al., 2011). For example, Lindgren et al. (1954) tested ten fumigants against stored-product insects and found that, in general, the 97 resistance (LD95, 6 hr exposure) of the stored-product pests was T. confusum > S. oryzae > R. dominica. In another study, S. oryzae generally had a higher percent mortality at lower doses than did T. confusum for many insecticides (Strong and Sbur, 1961). In general it has been noted that T. castaneum and T. confusum are harder to kill than other stored-product beetles (Arthur and Subramanyam, 2012). In another study, Bernays et al., (2000) showed that feeding by a generalist caterpillar (Heliothis virescens) was less strongly deterred by plant secondary metabolites than was the feeding of a congeneric caterpillar (H. subflexa). These data are all consistent with specialist insects showing greater sensitivity to compounds designed to result in mortality, and based on my study, it also appears that the secondary pests show greater sensitivity to compounds that result in a reduction in feeding behavior. Bernays et al. (2000) proposed that there is likely a trade-off to these two strategies. A specialists increased deterrence provides greater protection for that insect but limits the host range, whereas the reduced deterrence for the generalist allows it to have a wide host range and therefore more opportunities but an increased risk of negative impacts. It is important to note that our understanding of the relationship between these species of insects and their hosts prior to human-created envirnoments is limited. Furthering our understanding of this could help in the development of control methods and may help explain some of the differences that have been observed in sensitivity to chemical controls both in this study and in others. 98 Feeding deterrence and toxicity One of the most effective compounds against all the species except T. confusum (which was not deterred from feeding by any of the compounds tested including DEET) was 3c (4,n5). There was no significant difference when the butyl substitution was changed to either another pentyl or to an allyl functional group. However, for T. castaneum, shortening the carbon chains of side groups to either an ethyl or propyl moiety reduced the deterrent effect on feeding. Sitophilus zeamais and S. oryzae showed less sensitivity to shortening the substitutions than T. castaneum did. Rhyzopertha dominica showed feeding deterrence to the smallest compound which declined with larger substitutions until pentyl substitutions were reached. It is likely that the length of the substitution had an effect on how well the molecule interacted with components of the peripheral chemosensory systems, as described earlier, which changed the signal that was sent to the insect’s brain. Compounds with similar structure can show similar biological effects but do not necesarily show enhanced effects when combined. Three very similarly structured compounds isolated from the neem tree (salannin, salannol, 3-O-acetyl salannol), which showed antifeedant properties without mortality on lepidopteran larvae, showed no increase in effect when combined (Koul et al., 2004), unlike the enhanced activity observed with non-azadirachtin limonoids that are structurally different (Koul et al., 2003). Thus, structurally similar compounds likely compete for the same target site and show the same mode of action (feeding deterence) (Koul et al., 2003; Koul et al., 2004). My test compounds were structurally very similar, but despite this I found different modes of action within the same species (antifeedants vs. mortality), such as 3c{2,2} (reducing feeding with no mortality) vs. 3c{4,n5} (reducing feeding with high mortality) for R. 99 dominica. This would be an indication that these very structurally similar compounds were not competing for the same binding site which in turn could suggest that that the binding sites targeted by these compounds are highly specific, with even the smallest structural change having a large effect on the molecules’ relationship with a binding site and affecting the activity of the compounds. Understanding how these compounds may be binding and to what types of receptors would help clarify why and how they work. Substantial work with lepidopteran pheromones and their pheromone binding proteins and receptors has shed some light on how some receptors work (Prestwich et al., 1995; Honson et al., 2003; Mohanty et al., 2004; Wanner et al., 2010). Receptors have been found that were both extremely sensitive to different isomers of the same pheromone compound and some that were not (Wanner et al., 2010). Pheromone binding protiens in the same insect have shown binding affinities for different pheromone components, further indicating that these molecules play a role in differentiating between pheromone chain length (Prestwich et al., 1995). Changing chain lengths in pheromone analogues of Noctuidae species from the optimum length significantly reduced electroantennogram responses (Priesner et al., 1975). The small structural differences in my library of test compounds may result in failure to bind sucessfully with the structure of the binding proteins and/or the receptors. There is also evidence that the pheromone and the pheromone binding protiens interact together to activiate the receptors (Forstner et al., 2009); therefore, failure by a compound to successfully bind with an appropriate binding protein could result in the receptor not being activated or activated weakly. The active 3c compounds could also dock in the ligand binding domains in variant iontopic receptors (Dr. Erika Plettner, SFU, personal communication). 100 Variant iontopic receptors (IRs) are a second set of odor and taste receptors in arthropods (Benton et al., 2009; Rytz et al., 2013; Guo et al., 2014). They are derived from glutamate receptors with an altered binding site that binds hydrophobic odorants instead of glutamate. Insects also have odorant receptors (ORs), with seven transmembrane regions, that pair with a co-receptor (Orco) to be functional (Stengl and Funk, 2013; Larsson et al., 2004). Both IRs and ORs are involved in olfaction in insects and both cause depolarization of chemosensory neurons upon ligand activation which causes the chemosensory neuron to change its pattern of action potentials, an event that is detected in the deutocerebrum in the brain (Mori et al., 1999; Couto et al., 2005). Further studies to determine how the compounds may be binding may clarify how they are resulting in bioactivity. Only R. dominica showed reduced feeding in response to 3c{2,2}. This compound has the shortest alkyl groups tested and had no effect on either feeding or mortality for any of the other species. It also had no effect on mortality for R. dominica which means it has potential as a successful antifeedant compound for this species. With little to no mortality observed in the other species, it could be an indication that it would have relativly few non-target effects on other insects, and with very little mortality, the selective pressure which can result in resistance could be reduced. Substantial further testing would have to be done to determine if this is the case. Rhyzopertha dominica has been shown to be prevented from entering packaging by rotenone and helenalin; both are larger compounds than the less-effective juglone and geigerynin (Bloszyk et al., 1990). Rotenone is substituted with ethyl groups, as is 3c{2,2}, which may indicate that ethyl groups are an important functional group for feeding reduction of R. dominica, although helenalin does not have ethyl groups. This compound showed high mortality effects on 101 Trichoplusia ni in leaf disk choice bioassays (Akhtar et al., 2007), and twitching in the larvae was observed (Dr. Erika Plettner, SFU, personal communication). This indicates that for that species, 3c{2,2} may be affecting the central nervous system. Many pesticides, including pyrethrins to which R. dominica is particularly susceptible (White and Leesch, 1996), affect the nervous system by increasing sodium influx in voltage-gated sodium channels as well as interacting with chloride channels (Keifer and Firestone, 2007). However, 3c{2,2} did not have any effect on the mortality of the beetle species I tested, nor did I observe any twitching effects. During the experiment, I observed that T. confusum frass in several of the treatments was a dark color. On the control plates and a few of the treatments, T. confusum frass was very light colored, similar in color to the frass produced by T. castaneum. The insects feeding on both doses of 3c{3,3}, 3c{4,4}, 3c{3,6}, 3c{5,6}, 3c{5,5}, 3c{4,5} produced dark frass. These compounds caused increased mortality in S. oryzae, S. zeamais, and R. dominica; and that list includes the compound that caused mortality in T. castaneum (3c{5,6}). The difference in frass color may indicate that T. confusum is able to break down these compounds, resulting in a lack of bioactivity for this species. This ability may explain why this species showed no mortality or feeding deterrence regardless of the treatment. Analyzing the frass for the components that could be produced through the breakdown of the test compounds may provide further information about how the insects might be metabolizing the compounds. In addition, the levels of cytochrome P450s could be measured in the insects after consumption of the compounds. An increase in the level could indicate that the insects are using this mechanism to metabolize the compounds into non-toxic components. 102 As the insects were given three days to feed before any measurements were taken, the compounds may have caused reduced feeding due to either olfactory or gustatory signals. The choice of food is primarily attributed to contact chemoreception and avoidance is considered the outcome of chemoreceptors that often have a broad sensitivity spectrum to deterrents (Koul, 2008). However, the mode of action for chemicals that change feeding behavior is largely unknown (Koul, 2008). In bioassays of lepidopteran larvae with chalcones, flavones, and flavanone (also phenols), antifeedant activity was attributed to stimulation of deterrent neurons (Simmonds et al., 1990). The antifeedant activity is likely due to more than one receptor and the different receptor types probably have different structure-function type responses (Simmonds et al., 1990; Koul, 2008). It is also possible that feeding deterrents not only stimulate deterrent receptors but could suppress other receptors that might send signals to feed. For example, azadirachtin stimulates deterrent receptors but also appears to suppress sugar receptors (Schoonhoven, 1988 as cited by Koul, 2004). Alternatively, it is possible that insects consume a small amount of the compound resulting in a sub-lethal toxic effect. A sub-lethal toxic effect can reduce feeding because of endogenous signals due to a physiological response, rather than because they are receiving gustory signals that discourage feeding. If the insect immediately rejected the food, that would be an indication that the compound was detected by chemoreceptors (probably on the mouth or antennae) and the stimulus indicated that the food was unacceptable. However, because I allowed the insect to settle and remain exposed to the treatments for three days before any disturbance/observation occurred for the sake of the overall feeding assay, I was not able to make that type of observation. I did find a strong positive correlation between mortality and feeding for S. oryzae, S. zeamais, and R. dominica (Figures 4.2-4.6), suggesting that a lethal effect may have 103 contributed to feeding reduction for some of the species and compounds tested. This is not what was observed in feeding bioassays using T. ni where there was little correlation between feeding deterrent effects of these compounds and toxic properties (Akhtar et al., 2007). Many insecticides have an effect on the nervous system of insects and sub-lethal doses decrease feeding in some insects or negatively affect the insects’ ability to select hosts (Haynes, 1988; Fischer et al., 2014). As many of the test compounds that showed feeding deterrence also showed a lower LT50 than the controls and there was a significant positive correlation between feeding deterrence and mortality for many of the species, a sub-lethal toxic effect is a possible explanation. There is evidence that some insects have a post-ingestive response to consuming something that is not acceptable which in turn reduces feeding (Glendinnig, 1996; Glendinng and Slansky Jr., 1994). Glendinning (1996) showed that caterpillars rejected food with nicotine even after chemosensilla ablation within 30 seconds of feeding. Sub-lethal effects can be an effective tool for pest management (Foster and Harris, 1997). Because, 3c{4,4} reduced the feeding of T. castaneum and 3c{2,2} reduced the feeding by R. dominica without resulting in a LT50 that was significantly different than feeding on the control disks, for these two species and compounds it is unlikely that the feeding deterrence is a result of a sub-lethal toxicity. Therefore, it is more likely that these compounds interact with receptors which signal that the food is not acceptable. Alternatively, the compounds could have had a physiological effect on the insect that reduced feeding but ultimately did not result in mortality, perhaps due to an induction of detoxifying enzymes. 104 Summary I found several aromatic compounds which showed feeding deterrence without toxicity (3c{2,2} and 3c{4,4}), and could prove to be an effective tool in protecting stored products, although the effect was not seen for all the species tested. Those compounds could be effective in management tactics such as in conjunction with attractants in a push-pull strategy (Pyke et al., 1987; Cook et al., 2007). The compounds that deterred the insects from feeding could be used as the “push” in this type of pest management strategy. If the compounds show repellent effects, they could have potential to be used in packaging (Mullen et al., 2012). Insects can penetrate packaging, or invade using already existing holes (Highland, 1984). Repellent compounds such as DEET prevent stored-product insects including S. oryzae and T. castaneum from entering envelopes containing wheat (Hou et al., 2004). Other insecticides reduce penetration into polyethylene and jute bags (Abdelghany et al., 2016). I also have shown that there are some structure-activity relationships between the feeding deterrent effects which could allow further optimization of compounds or mixtures that create a desirable behavioral effect. Mixtures would likely reduce the development of resistance and habituation. As I observed very different behavioral responses by different species, even closely related ones, mixtures would have the potential to widen the group of insects affected (Koul et al., 2004). Of course, the individual components of the mixture must be thoroughly studied and tested. A pentyl substitution showed effective feeding deterrence across four of the storedproduct pest species tested although it also had an insecticidal effect. Behavioral activity observed in the laboratory is a substantial step in determining if there is any potential in developing a commercial product. Similar to insecticides, it is important for 105 behavioral-modification chemicals to have residual activity only as long as needed, to only cause below-threshold levels of product contamination, and be non-toxic to non-target organisms, including humans (White and Leesch, 1996; Isman, 2002). These and other factors would have to be determined in a regulatory process before any of these compounds could be successfully used alone or in combination with other tactics in an overall stored-product pest management strategy. 106 Chapter 5. A comparison of walking behavior of Tribolium castaneum and Sitophilus oryzae in response to seven benzene ring-containing compounds Abstract I tested two species of stored-product pests, Tribolium castaneum and Sitophilus oryzae, for behaviors related to avoidance of seven benzene ring-containing compounds, including DEET (N,N-diethyl meta-toluamide), various para-substituted dialkoxybenzenes and an N- and O-alkylated p-hydroxyaniline, using a walking bioassay. Assays were conducted using a filter paper treated on one half with a test compound dissolved in methanol at a rate of 5.2 µmol of compound per assay. Individual insects were allowed to walk on the filter paper and their location (treated or untreated side) was recorded every fifteen seconds for five minutes and then after fifteen minutes. There were no significant differences between treatments for either species in the percentage of time spent on either side of the filter paper. Sitophilus oryzae crossed the middle of the filter paper less often on the DEET treated paper and was found after fifteen minutes to be on the untreated control side of both the DEET and 3c{N3,O2} (N-propyl-4ethoxyaniline) treatments more frequently, which indicates avoidance of these compounds by this species. Further testing to determine if there are potential long- and short-range behavioral effects is necessary. Introduction One way to protect stored products from damage by a wide range of stored-product insect pests is to use chemicals (both natural and synthetic) to deter them. Manipulation of an insect pest’s behavior can protect resources (Foster and Harris, 1997) and many forms of this type of control involve the use of chemical stimuli, although other stimuli such as visual cues also can be 107 used (Foster and Harris, 1997). Stimuli can be described by their effect on the insect’s behavior such as attractants and repellents, which cause the insect to make oriented movements either towards or away from the stimulus (Dethier et al., 1960). One method that uses attractants to control insect populations is the ‘attract-annihilate’ or ‘lure and kill’ method. This method attracts the pest to a location where as many individuals as possible can then be removed not by just trapping them but with a killing agent (Foster and Harris, 1997; El-Sayed et al., 2009). This attract-annihilate method of control has been shown to have potential with Plodia interpunctella, Indianmeal moth (Lepidoptera: Pyralidae), using the insect’s pheromone and an insecticide (Campos and Phillips, 2014). In other instances pheromones, other semiochemicals, and light have been used to draw insects into traps away from the resource that needs to be protected (Levinson and Mori, 1983; Lindgren et al., 1985; Mullen, 1994; Stejskal, 1995; Watson and Barson, 1996; Duehl et al., 2011; Jeon et al., 2012; Mullen et al., 2012). Attract-and-kill control could also be enhanced with a push-pull strategy where a stimulus “pushes” the insect away to protect the resource and another stimulus “pulls” the insect to a different location (Pyke et al. 1987; Cook et al., 2007). Trapping can serve as both a way to reduce pest populations (control method) and as a way to monitor the numbers in a pest population. Monitoring stored-pest populations is essential to determine the economic threshold at which control measures need to be taken, and is an essential part of an integrated pest management program (Hagstrum and Subramanyam, 2012). After all, there are costs associated with the use of insect control methods, and inappropriate use of insecticides can increase the negative consequences associated with them (Devine and Furlong, 2007). 108 While previous bioassays showed that some phenol-derived test compounds reduced feeding and/or resulted in toxicity to several species of stored-product insects (Chapters 2-4), I was unable to identify the mechanism of deterrence based on those results alone. The insects may have avoided the treated food disks based upon the volatiles emanating from the compound in the disk or may have been influenced by their tactile or gustatory response to the disks. The volatility (LogK1/g) was determined for several of the compounds tested (Ebrahimi et al., 2013) although more research is being done on the physical properties of these test compounds. An alternative explanation is that the compounds have a systemic effect on the insect that slowed down their movements and/or consumption of the food. Therefore, I tested a limited number of the compounds with two important species of stored-product insects: the red flour beetle, Tribolium castaneum (Coleoptera: Tenebrionidae), and the rice weevil, Sitophilus oryzae (Coleoptera: Curculionidae), using a walking bioassay to test if the compounds had a repellant effect either due to an olfactory or tactile responses rather than a postingestive effect. The compounds used in this study (Table 5.1) showed a range of activity in previous feeding bioassays (Chapter 3 and 4). Materials and Methods Insects All insects used for the walking bioassay were from the same source as those used in the feeding bioassays (Chapter 4). Tribolium castaneum and S. oryzae adults were from laboratory cultures kept at 30oC, 70% r.h. for over five years, and reared on wheat flour mixed with 5% brewer’s yeast and wheat kernels respectively. Prior to use in the experiment, insects were removed from their food medium and stored in a vial for up to two hours. Males and females 109 were kept separately at this point, although their mating status was unknown. Only adults that appeared visually healthy (e.g., motile, all limbs present) were used. Treatments Seven test compounds were used during the walking bioassay along with a no-treatment control and a carrier solution (HPLC-grade methanol) control (Table 5.1). Included in the table are the responses by both species tested to the compound in previous feeding bioassays. With the exception of 3c{N2,O3} all behavioral results are from one bioassay (Chapter 4). As 3c{N2,O3} was not used in that bioassay, the result presented is from a separate feeding bioassay (Chapter 3) in which only S. oryzae was used. The compounds were created using the methods noted previously (Chapter 2 and 3). Compounds were dissolved in HPLC-grade methanol to create a dose of 5.2 µmol of compound per half filter paper (5.2 µmol /2.7 cm2) which is the equivalent concentration of compound found in the high dose of one flour disk from the previous feeding bioassay (see Chapter 3). Walking bioassay The walking bioassays were conducted at the base of a glass vial (2.6 cm diameter) in the middle of a fume hood with equivalent lighting on both sides. A thin line of pencil was drawn bisecting the filter paper. One half of the filter paper was left untreated, the other half was treated using 40 µL of the treatment compound. For control assays, no liquid was added to either side. For the methanol control (the solvent used), only methanol was added. The treated filter paper was allowed to dry in the fume hood for several minutes so that it was not visibly wet when the insect was introduced. The filter paper was placed inside the glass vial and the vial was treated 110 Table 5.1. Structures of test compounds used in walking bioassay of Tribolium castaneum and Sitophilus oryzae. The varying alkyl substituted groups are listed. The response for both species tested is from the previous feeding bioassays (Chapter 3 and 4). Minus indicates significantly (P < 0.05) decreased percent feeding and NS indicates no effect on feeding after three days at 26 µmol per replicate compared to the control flour disks. Response Treatment R1 R2 Compound T. castaneum S. oryzae 3c{3,3} Propyl Propyl 1,4-dipropoxybenzene NS NS 3c{3,6} Propyl Allyl 1-(allyloxy)-4-propoxybenzene NS NS 3c{4,4} n-Butyl n-Butyl 1,4-dibutoxybenzene – – 3c{4,n5} n-Butyl n-Pentyl 1-butoxy-4-(pentyloxy)benzene NS – 3c{n5,n5} n-Pentyl n-Pentyl 1,4-bis(pentyloxy)benzene NS NS 3c{N2,O3} Ethyl Propyl N-ethyl-4-propoxyaniline DEET N/A N/A N,N-diethyl meta-toluamide Control N/A N/A MeOH Control N/A N/A –* NS – *Tested in a different bioassay from the others (Chapter 3) and only on S. oryzae. 111 around the base with polytetrafluroethylene to prevent the insects from climbing out of the arena. A single insect was placed at the edge of the vial along the dividing line, and was observed walking around on the disk for five minutes. Every 15 seconds the insect was recorded as being on the treated or untreated side of the filter paper. During the entire time the insect was in the arena, the number of times the insect crossed the treatment line was observed to serve as a measure of insect activity. After the initial five-minute observation period, the insect was allowed to remain in the vial and then was observed at 15 minutes to see which side of the vial it had settled on. After the experiment was complete, the insect was removed and a new insect of the same species and same sex was added to the filter paper and the same observations were taken to maximize the information that could be obtained from the test compounds. Between each test the side that the treatment faced was switched to try to account for any external influences, such as light or temperature variation, that may have an effect on the behavior of the insect. Each treatment was done with eight insects, four male and four female with the exception of DEET for T. castaneum which only had four treatments in total. There was a limited amount of the test compounds available which limited the number of replicates. Analysis. The methanol control tests were analyzed using either a t-test or a Mann-Whitney U test, based on whether the data met requirements of normality and equal variance, to see if the side the treatment faced (either left or right) had an effect on the proportion of time the insect spent on it. The proportion of time each insect spent on either the treated or untreated side of the filter paper based on the 15-second observations was calculated. Data were analyzed using a Bonferroni-adjusted chi-square for multiple proportions. 112 The number of times each insect crossed the middle line of the area during the five minutes of observation was analyzed using ANOVA followed by a Tukey’s HSD post-hoc test when a significant difference was detected to compare treatments. To compare the number of times the insects crossed the middle line between the two species, only the controls were analyzed using a t-test as the data met the requirements of normality (Shapiro-Wilk test) and equal variance. The proportion of the insects found on the treated or untreated side after fifteen minutes was analyzed using z-tests. All statistical analyses were done using Sigma Plot 12.5 except for the Bonferroni-adjusted chi-square for multiple proportions, which was calculated by hand. Results Treatment location There was no difference in the proportion of time either species spent on the treated or untreated side of the methanol control based on which side of the fume hood the treatment faced (U = 10.0, P = 0.69). Therefore, the side of the fume hood the treatment faced was not considered in subsequent analyses. Proportion of time There were no significant differences (P > 0.05) between any of the treatments in the walking bioassay for the proportion of time spent on either the treated or untreated side of the filter paper during the first five minutes for either T. castaneum or S. oryzae (Figure 5.1). The lowest percentage of time S. oryzae and T. castaneum spent on any treatment was on DEET. 113 Figure 5.1. Mean percentage of time spent on either the treated or untreated side of the filter paper based on 15-second observations over five minutes (±SE) by Tribolium castaneum and Sitophilus oryzae. There were no significant differences between any of the treatments within species using a Bonferroni-adjusted chi square for multiple proportions. 114 Crossing There were no significant differences in the number of times T. castaneum crossed the middle line of the arena during the first five minutes in any of the treatments (F8,61=1.87, P = 0.082) (Table 5.2). There were significant differences in the number of times that S. oryzae crossed the middle line (F8,63=6.42, P < 0.001) (Table 5.2). DEET resulted in S. oryzae crossing the least number of times while 3c{n5,n5} resulted in the highest average number of crossings although neither was different from either of the controls. In the control assays, T. castaneum crossed the middle line of the arena significantly more often compared with S. oryzae (t14 = -3.4, P = 0.004). Fifteen minute location There was no significant difference between the treatments and the location the T. castaneum were found at after being allowed to remain in the arena for fifteen minutes (P > 0.05); however, there was a significant difference between treatments for the S. oryzae (P < 0.05) (Table 5.3). The insects were found on the side treated with DEET and 3c{N3,O2} the least often although it was not significantly different than the control. There also appears to be a slight attraction (again non-significant) by S. oryzae to 3c{3,3} and 3c{3,6}. 115 Table 5.2. Mean number of times (±SE) individual insects crossed the dividing line in the middle of the filter paper over the five minutes for both Tribolium castaneum and Sitophilus oryzae. Half of the filter paper was treated with one of the test compounds. Significant differences (P < 0.05) determined using analysis of variance followed by Tukey’s HSD test (n=8) are represented by different lowercase letters in the S. oryzae column. Treatment Mean number of times the dividing line was crossed (±SE) T. castaneum S. oryzae 3c{3,3} 17.6 (± 2.1) 10.5 (± 2.1)ab 3c{3,6} 19.4 (± 2.9) 5.6 (± 1.1)bc 3c{4,4} 12.5 (± 1.4) 9.1 (± 1.7) abc 3c{4,n5} 9.6 (± 1.3) 10.6 (± 0.8)ab 3c{n5,n5} 14.8 (± 1.7) 15.8 (± 1.7)a 3c{N2,O3} 14.8 (± 2.8) 5.3 (± 1.0)bc DEET 15.0 (± 4.1)* 2.5 (± 0.6)c Control 18.9 (± 2.0) 11.3 (± 2.7)abc MeOH Control 15.3 (± 1.0) 9.6 (± 1.0)abc *n=4 116 Table 5.3. Percentage of insects found on the treated side of the filter paper after fifteen minutes for each species. Significant differences in columns represented by lowercase letters (n=8). Species Treatment T. castaneum S. oryzae 3c{3,3} 50.0 87.5a 3c{3,6} 50.0 87.5a 3c{4,4} 37.5 37.5ab 3c{4, n5} 12.5 62.5ab 3c{n5,n5} 62.5 62.5ab 3c{N2,O3} 50.0 12.5b DEET 0.00* 12.5b MeOH control 50.00 37.5ab *n = 4 117 Discussion Based on these walking bioassays, T. castaneum did not show any behaviors that would indicate avoidance of the treatments, including DEET. The lack of response by the red flour beetle is consistent with the lower response to these compounds that I saw in the feeding bioassay (see Chapter 3). Tribolium castaneum moved around the arena more than S. oryzae as evidenced by the number of times the insects crossed the middle line of the arena on untreated filter paper as a measure of comparative activity between the species. This high level of movement could have hidden any mild avoidance as the insect may have been more motivated to keep moving (and therefore keep being observed on the treated area). Sitophilus oryzae exhibited a reduced number crossings of the middle line into the DEET treated area, although it was not significantly different from the control. DEET was also the treatment on which the insects spent the lowest percentage of time (n.s.) as well as the treatment the insects were less likely to be found on after fifteen minutes (not different from control). Together this suggests that S. oryzae avoids contact with DEET which is consistent with Hou et al.’s (2004) findings that DEET was very effective at preventing insects, including S. oryzae, from entering into envelopes as it is a known repellent against several stored-product insects (Khan and Wohlgemuth, 1980; Watson and Barson, 1996). Test compound 3c{N3,O2} was the next most effective for S. oryzae in all three of the same tests. Again, while not significantly different from the controls, it indicates that further testing of this compound may show deterrent effects that are not related to feeding. This and other nitrogen-substituted compounds showed a strong feeding deterrence effect on S. oryzae and also resulted in high mortality (Chapter 3). As the compound resulted in mortality, avoidance of 118 the treated side of the filter paper could increase the chance of survival. I cannot determine from my data if 3c{N2,O3} has any contact toxicity or if the compound must be ingested to be lethal. 3c{N2,O3} was the one test compound which also contains an aromatic ring substituted with nitrogen, which is similar to DEET. DEET shows lethal effects on S. oryzae and in these bioassays the insects also exhibited some measure of avoidance of DEET, similar to that observed with 3c{N2,O3}. DEET may work by blocking insect odorant receptors that would respond to attractive food (Ditzen et al., 2008), and there is evidence that DEET is an odorant with repellent properties (Sfara et al., 2011). However, the mode of action of DEET may also be to target octopaminergic (insect homolog of adrenaline) synapses resulting in insect mortality (Swale et al., 2014). It is possible that 3c{N2,O3}’s similar structure to DEET imparts it with a similar mode of action. There are several limitations to this study and the results should be interpreted with caution. I chose to use a small arena which ensured the insect came into contact with the compounds and allowed a very small amount of compound to be used. However, the small arena size could also have resulted in a saturation of the insects’ receptors preventing the insects from effectively making a choice. The small arena could also have resulted in the insects being unable to successfully differentiate between the treated and untreated side of the filter paper. Further, the exposed arena could also have an influence on the insects’ behavior as both of these species show responses to light (Arbogast and Flaherty, 1973; Duehl et al., 2011; Jeon et al., 2012). While I rotated the vials to control for a left/right bias, the insects were exposed to light during the entire test while they would naturally live in a dark environment. It has been observed that T. castaneum under high light intensities moved less and tried to hide (Semeao et al., 2011). The insects could have been more motivated to try and find a hiding place due to the light stimulation 119 and therefore disregarded any other stimulus including the test compounds. Effects of combined sensory inputs in insects has been well-documented – e.g., light and olfactory cues (Duehl et al., 2011; Semeao et al., 2011) – and simultaneous sensory inputs can conflict with one another. I also only tested the compounds at one dose level due the limited amount of test compounds available, and it is possible that response could be dose-dependent. My tested concentration of DEET is similar to the low concentrations tested by Khan and Wohlgemuth (1980) which showed a mild attraction by T. castaneum (with repellent effects at higher doses), though I did not detect that effect. My dosage of DEET is also similar to that used by Hou et al. (2004) who observed a reduction in packaging penetration by both S. oryzae and T. castaneum; but the dose was lower than that used in other experiments with stored-product insects (Anthrenus verbasci, Oryzaephilus surinamensis) (Watson and Barson, 1996; Watson et al., 1997). It is also possible that because more than one insect was used in the arena, the first insect may have had an effect on the second. Further, I only tested a small number of insects which limited my ability to detect behavioral choices being made. More significant results may become more apparent with an increased number of replicates. Despite these limitations, my results indicate that at least one of the compounds tested – 3c{N3,O2} – may have behavioral effects on S. oryzae which could be studied further. If these compounds or others like them were to be shown to be effective deterrents or locomotive initiators, they could prove useful in a stored pest management strategy. This would require extensive additional testing including determining vertebrate toxicity, persistence, environmental effects, and many others. However based on the initial work, this may be worth further study. 120 Chapter 6. An inexpensive feeding bioassay technique for stored-product insects This chapter has been previously published: Clark, E.L., Isitt, R., Plettner, E., Fields, P.G., and D.P.W. Huber (2014). An inexpensive feeding bioassay technique for stored-product insects. Journal of Economic Entomology 107: 455-461. DOI: http://dx.doi.org/10.1603/EC13283 Abstract I used the red flour beetle, Tribolium castaneum (Coleoptera: Tenebrionidae), to compare three feeding bioassay techniques using flour disks. The area (scanner or digital photographs) and mass (sensitive balance) of the same flour disks were measured daily for one or two weeks to assess feeding by insects. The loss in mass and area over four hours was measured, as some variation over time was noticed in the disks with no insects feeding on them. The gravimetric method correlated well with both measurements of the area for the disks held in a growth chamber: scanner (R2=0.96), digital photography (R2=0.96). There was also a high correlation (R2=0.86) between the disk weight and area scanned at normal lab conditions. There were differences in the percentage of the disks remaining over time depending on the temperature and whether they were weighed or scanned. Measuring the mass of the disks resulted in a relatively larger percent of disk remaining compared with the scanned area. Mass measurements required a sensitive balance, handling of the disks and the insects, and appeared slightly more sensitive to humidity and temperature changes over time. Scanning the disks requires flat bed scanner access but less handling of both insects and disks. Digital photographs could be taken quickly, requiring less equipment, although photographs had to be further processed to determine area. Scanning or 121 taking digital photographs of flour disk area was an effective technique for measuring insect feeding. Introduction A wide range of bioassay techniques have been developed for testing the effects of natural and synthetic compounds on the feeding of insects (Koul 2004). Techniques for measuring insect feeding vary depending on factors such as the type of compound being tested, the life stage of the insect, and the type of food. Therefore, the ways in which feeding activity is measured in different situations also varies. For example, a classic method of measuring feeding activity for phytophagous insects uses uniform disks punched out of leaves (e.g., Wijkamp and Peters, 1993; Koul, 2004). Following feeding on treated or control leaf disks, the remaining leaf area can be measured and compared with other treatments. The red flour beetle, Tribolium castaneum Herbst (Coleoptera: Tenebrionidae), is a pest of stored grain products and is found worldwide. Xie et al., (1996) developed a gravimetric feeding bioassay that is an effective method of testing a wide range of potential deterrent or toxic compounds (e.g. Liu and Ho 1999, Huang et al., 2000, Hou et al., 2004, Fields et al., 2010). In this method, the flour disks are weighed to determine differences in the amount of feeding between treatments over time. It requires a sensitive balance - one that can measure as small a difference as 0.1 mg. In this chapter, I compare the gravimetric disk bioassay (Xie et al., 1996) with the disk area eaten, a technique widely used in leaf consumption bioassays (e.g. O’Neal et al., 2002). I conducted feeding bioassays using red flour beetles in which I measured the flour disks using three different methods: scanning to record surface area, using digital photographs to determine surface area, and weighing the disks. 122 Methods Experimental insects Tribolium castaneum were reared on organic whole wheat flour with 5% b.w. brewer’s yeast at 30oC. The adult insects used in the experiment were 0-21 days old at the beginning of the experiment. Adults that were visibly healthy were removed from the flour jars the day before the experimental trial and were starved overnight before the beginning of the bioassay. Disk manufacture Flour disks were made using slightly modified methods from Xie et al., (1996). Unbleached organic white flour (800 mg) and distilled water (4 mL) were stirred using a magnetic stir bar for a minimum of two minutes. Aliquots (100 µL) were pipetted onto aluminum weigh boats (Fisherbrand, Fisher, Pittsburgh, PA). The weigh boats were then covered with plastic Petri dishes (Fisherbrand, Fisher), allowing slight airflow, and allowed to dry at room temperature overnight. The following day the disks were put into the plastic Petri dishes and placed into a covered plastic bin with a beaker of 200 mL of saturated NaCl aqueous solution to allow stabilization of moisture content for 24 h. Methods for estimating feeding: Scanning vs. weighing To compare scanning and weighing, the flour disks were removed from the bin after equilibrating for 24 h. Bioassays were conducted keeping the disks in a bin at room temperature (26 ± 3oC, 68 ± 8% RH in the bin) and in a growth chamber (29 ± 2oC, 74 ± 10% RH in the bin). For both room temperature and growth chamber sets, five disks were placed in each of 25 plastic Petri dishes, before flour beetles were introduced. The flour disks were scanned (Epson Expression 1640XL, Long Beach, CA) to calculate the surface area of the disks (WinFOLIA Pro 123 2003d) as well as weighed (SI-235 analytical balance, Denver Instrument, Arvada, CO, measures to 0.1mg) for initial measurements. Twenty-five T. castaneum adults that had been removed from the colony the previous day were added to each Petri dish and were placed in a covered plastic bin (37.4 x 24.1 x 14 cm) lined with aluminum foil to keep light out. In addition, ten plates of five flour disks containing no beetles (non-feeding control) and five plates containing no flour disks, but 25 insects (starved control) were also placed in each bin. A beaker with 200 mL of saturated salt (NaCl) water solution was placed in each bin to maintain humidity. A temperature and humidity gauge (Accu-temp Digital Thermometer, Springfield Instrument Company, Montreal, QB) was also placed in each bin. The covered bins were sealed using duct tape to keep the humidity high. One bin (40 Petri dishes) was left at room temperature while the other (40 Petri dishes) was placed in a growth chamber at 30oC. All disks were weighed and scanned every day for two weeks and beetle mortality was noted. Temperature and humidity were recorded immediately on opening the plastic bin. I alternated which disks were weighed first while the other disks were simultaneously scanned (room temperature vs. growth chamber). Photographing vs. weighing To compare digital photography and weighing to estimate feeding, the disks were prepared and humidified as described in the Disk Manufacture section. However, only ten plates with insects and five plates without insects (no feeding) were kept in a bin at 30oC and measurements were only taken over one week. The photographs were taken using a Sony CyberShot digital camera (DSC-H1, 5.1 megapixels) set on a tripod to ensure that the camera was the same distance from the disks every time (Figure 6.1C). A piece of cardboard was used to prevent glare from the Petri dish and a black background was used as contrast to the flour disks. Before 124 photographing disks, a photograph of metric graph paper was taken and used to calibrate the number of pixels in a square centimeter. Disks were removed from the plates to separate them from the insects and then weighed and photographed immediately. Obtaining the mass and the photographs for all the disks took 20 minutes. After taking the photographs, disk area was then calculated using the open-source GNU Image Manipulation Program 2.6.11 for Windows (GIMP Development Team 2010). Using the contrast of the flour disks against the dark background allowed the number of pixels in the disks to be determined (Figure 6.1A). The area of the flour disks was calculated by using the number of pixels in the known area on the graph paper. Loss on bench top Based on preliminary results, there was a noticeable reduction in recorded mass and recorded area using scanning over approximately four hours. To determine if the length of time that the disks sat at room temperature and how humidity affected the mass and area of the flour disks I made new flour disks as described previously. However, no beetles were added to the disks. The disks were humidified for a minimum of 24 hours either at room temperature or at 30oC before being removed from their bins. The disks were then scanned and weighed in less than five minutes after the bin was initially opened. The flour disks being weighed and scanned were kept at room temperature and humidity on the bench after the initial mass and area were recorded (21.1oC, 20% RH) and then re-weighed and re-scanned every hour for four hours. Data analysis: Scanning vs. weighing The number of dead beetles after 14 days was analyzed using Wilcoxon rank sum tests, as the data did not meet requirements for a parametric analysis. The mortality between the room temperature and growth chamber starving controls and between the room temperature and 125 A C B D Figure 6.1. A: Photograph of disks. The top row has had no insect feeding, the bottom row has had one week of insect feeding. B: A scan of plates of disks. More than one plate can be scanned at a time. C: Set-up for taking photographs of the disks. D: Example of photograph of disks after one week feeding. 126 growth chamber feeding plates were compared. To ensure a balanced sample size, a randomly selected subset of five feeding plates was compared with mortality in the five starving controls at each corresponding test temperature. Feeding plates with fewer than 25 living flour beetles remaining after 14 days were removed from the analysis (five plates in the room temperature bin, two plates in the growth chamber). Three more plates from the growth chamber bin were randomly selected and excluded from analysis to ensure equal sample sizes. In order to compare measurements of feeding area (cm2) versus mass (g), the data for each plate were transformed into the proportion of flour disk remaining (relative to day 0). The data did not meet requirements of parametric analysis. Therefore, the proportion of each flour disk remaining between area and mass-based measures within the growth chamber bin were compared using Wilcoxon signed rank tests for each day. The flour disks held in the room temperature bin were analyzed in the same way. To compare the proportion of each flour disk remaining using area measurements between the growth chamber and room temperature bins for each day, Wilcoxon rank sum tests were used. The same method was used to compare the proportion of each flour disk remaining using mass measurements for each day between the two bins. A linear regression between the disk area (cm2) and the disk mass (g) of flour disks from both the room temperature and growth chamber was also created. Photographing vs. weighing A linear regression between the disk area (cm2) recorded using photographs and the disk mass (g) recorded using a scale was calculated, and did not include the no-feeding control plates. Loss on bench top The changes in both disk area and in mass while disks were kept on the lab bench were non-linear. A Michaelis-Menten model was used to describe the change in disk area (cm2) 127 recorded using a scanner and the change in disk mass (g) recorded using a scale over the fourhour time period. All data were analyzed using R v. 2.15.2 (R Core Team 2012). Results Scanning vs. weighing Both the mass and area declined due to feeding by T. castaneum. After 14 days only an average (± SE) of 15.41 ± 0.01% of the recorded mass and 14.61 ± 0.01% of the recorded area remained for the disks held in the growth chamber (Figure 6.2). Beetles held at room temperature left 44.12 ± 0.02% of the disk mass and 37.76 ± 0.02% of the disk area after 14 days. For all dates but one, the flour beetles held at 30oC ate more than the ones held at room temperature (Figure 6.2). The average percent remaining of the disks as measured by mass and by area using a scanner were significantly different (P < 0.05) between the growth chamber and room temperature bins for all days (Figure 6.2). The decline in area with time was greater than the decline in mass for both the growth chamber and the room temperature experiments (Figure 6.2). For example, the day following the initiation of the experiment at which 50% or less of the disk’s measured mass or area remained differed depending on the measurement technique and environmental conditions. In the growth chamber bin, on average less than 50% of the disks remained at day 8 if measured using mass, and at day 6 if measured using area. 128 Figure 6.2. A: Temperature (oC) and percent relative humidity inside the box when first opened on each day. B: Percent of disk remaining (mean ± SE) in each box (room temperature and growth chamber), measured by mass and area, fed on by 25 adult T. castaneum. There were significant differences (P < 0.05) in the percent of disk remaining from area and mass based data between the growth chamber bin flour disks and room temperature bin flour disks for all days. There were significant differences (P < 0.05) between the growth chamber bin percent of disk remaining from the area and mass measurements and between the room temperature bin percent of disk remaining area and mass measurements for all days. 129 There were no significant differences between the mortality of beetles held without food at room temperature (3.20 ± 0.49) and in the growth chamber (5.00 ± 1.00) (W = 19.5, P = 0.0156) or with food at room temperature (0.24 ± 0.12) and in the growth chamber (0.04 ± 0.04) plates (W = 274, P = 0.157) after two weeks. I observed higher mortality in the starving controls compared to the feeding plates for both room temperature (W = 0, P = 0.007) and growth chamber (W = 0, P = 0.007). There was a significant correlation (P < 0.001) between the disk mass (g) and the disk area (cm2) for both the disks held at room temperature and for the disks held at 30oC (Figure 6.3). The relationship between area (cm2) and the mass of the disks (g) is described by the following equation for the disks held at room temperature (Figure 6.3a): Area = 36.59 (±0.42) (mass of disk) – 0.07 (±0.02) (R2 = 0.86). The relationship between area (cm2) and the mass of the disks (g) is described by the following equation for the disks held in the growth chamber (Figure 6.3b): Area = 35.58 (±0.83) (mass of disk) – 0.14 (±0.05) (R2 = 0.96). Photographing vs. weighing There was a significant correlation (P < 0.001) between the disk mass (g) and the disk area (cm2) using the disk photographs over seven days (Figure 6.4). The relationship between area (cm2) and the mass of the disks (g) is described by the following equation: Area = 32.03 (±0.78) (mass of disk) – 0.21(±0.05) (R2 = 0.96). Loss on bench top The Michaelis-Menten model varied depending on the treatment (Table 6.1). The measurements using mass had a higher percent loss asymptote than the measurements using area. In all four treatments most of the loss in area and in mass occurred within the first few hours (Figure 6.5). 130 Figure 6.3. Linear regression (± 95% CI) of the flour disk area (cm2) based on scanning and mass (g) between the disks held at (A) room temperature and (B) held at 30oC. There was a significant correlation for both (P < 0.001) represented by the equations: (A) Area = 36.59 (±0.42) (mass of disk) – 0.07 (±0.02) (R2 = 0.86) and (B) Area = 35.58 (±0.83) (mass of disk) – 0.14 (±0.05) (R2 = 0.96). 131 Figure 6.4. Linear regression (± 95% CI) of the flour disk area (cm2) based on photographs and mass (g). All disks were held at 30oC. There is a significant correlation for the linear regression (P < 0.001) represented by the equation: Area = 32.03 (±0.78) (mass of disk) – 0.21(±0.05) (R2 = 0.96). 132 Figure 6.5. Mean percent loss (±SE) of flour disk area (cm2) and mass (g) over four hours at room temperature and humidity (21.1oC, 20% RH). 133 Table 6.1. The Michaelis-Menten equation Change (%) = [Vmax * Time (h)] / [Km + Time (h)] for the four treatments where Vmax is the asymptote and Km is the Michaelis constant which represents the length of time in which the disk would have lost half the water. Treatment Measurement Vmax (h) Km (h) Residual sum of squares Room Temperature Area -6.0 1.8 171.7 Room Temperature Mass -6.4 1.0 10.45 Growth Chamber Area -4.9 1.1 25.93 Growth Chamber Mass -6.5 1.2 22.61 134 Discussion The gravimetric method developed by Xie et al. (1996) has proven to be an effective tool for examining the effects of compounds on feeding and survival of stored-product insects (Liu and Ho 1999, Huang et al., 2000, Hou et al., 2004, Fields et al., 2010). In this chapter I present alternative methods using a scanner or a digital camera to calculate disk area. These inexpensive techniques correlate well with the gravimetric method that requires an expensive sensitive balance. The advantages of the gravimetric method are that the data can be compared with previous experiments (e.g. Lin and Ho 1999, Fields et al., 2010) and it requires little treatment of the data after the measurements are taken. The disadvantage is the necessity of a sensitive scale and both the disks and the insects must be handled which, in some cases, can result in small pieces of disks being difficult to weigh due to static buildup. The advantages to the scanned method are less handling of the disks and insects during the measurement and, depending on the program used to obtain the disk area, the area can be determined right away. The disadvantages are that the image is two-dimensional so if the disks curl, that may increase the variation in the measurements. Also, there can be small disk pieces after the insects have fed which may not be detected by the scanner. The advantages to the digital photographs are that they require very little equipment and the photos can be taken rapidly. The area analysis can also be done using opensource software. However, this method does require the area of the disks to be calculated from the photographs, which takes additional time and, as with the scanning method, the image is twodimensional meaning any curve in the disk will change the measured area. These differences may be important to consider depending on the specific requirements of an experiment such as sensitivity of insects to handling. 135 There was a loss of both mass and area when the disks sat out at room temperature for four hours even with no insects feeding (Figure 6.5). I predict that this is owing to a loss of moisture from the disks. This would affect the disk mass and it likely resulted in the disk edges curling slightly. This change in shape would have reduced the flour disk area that the scanner (taking two dimensional images) could detect. This is an important consideration if taking all measurements is likely to take a long period of time, particularly as about half of the loss occurred in the first couple of hours, although the loss I observed was generally under 5%. However, good planning should ameliorate this difficulty. For instance, the difference that I observed would likely vary depending on the conditions of the room in which the experiment is being conducted. The mass of the disks likely showed, on average, a higher percentage decrease than the area of the disks because the change in disk shape detected by the scanner was smaller than the loss of mass from the moisture. A higher percent loss was found at increased feeding temperatures, which is consistent with flour beetles developing more quickly at 30oC than at 25oC (Howe 1956). Although the percent loss due to feeding over time varied significantly depending on the way in which it was measured (area from a scanner vs. mass), there was significant correlation between the area measurements and the mass measurements. An expensive, sensitive balance - which is required for weighing the disks to an appropriate level of accuracy – may not be available in all situations, but either scanning or taking photographs can be used with equal accuracy. In addition, there are open-source software sources available for analyzing the images, maintaining the low cost of this method. A higher resolution camera would also increase the sensitivity of using digital photography to determine feeding. It is also worth noting that there was a strong correlation even if the disks were held at room temperature. This indicates that even if an incubator is not 136 available the technique could still be used. The significant correlation between the three measurement techniques makes them comparable with each other in terms of accuracy and allows researchers with limited resources to assess feeding and compare their results to those of others. 137 Chapter 7. Conclusions The development of new control tactics for stored-product insects is of high importance. Due to the damage that they cause to our food supply at every stage (Hagstrum and Subramanyam, 2006; Hagstrum et al., 2012), along with the high number of pest species that now show resistance to one or more insecticide (Fields, 1992; Champ and Dyte, 1977; Subramanyam and Hagstrum, 1995), and the interest in using behavioral control tactics that have fewer toxic effects (Subramanyam and Hagstrum, 2000), there is a need to develop novel tactics for use in control strategies. Based on numerous bioassays I was able to identify several diakloxybenzenes that resulted in feeding deterrence bioactivity in multiple species of storedproduct insects. Tribolium castaneum, an important pest of stored products and one that can be more difficult to kill than others (Arthur and Subramanyam, 2012), showed a large increase in mortality when feeding on one test compound (3c{3,6}) and reduced feeding on several other para-substituted aromatic rings (3c{4,4}, 3c{6,6}, 3c{3,n5}, 3c{n5,6}, 3c{4,6}, 3c{4,n5}, and 3c{n5,n5}). The feeding reduction was often observed without a significant increase in mortality (exceptions: 3c{3,6}, 3c{4,6}, 3c{6,6}). Using no-choice feeding bioassays I was able to screen a large number of compounds and reduce the number of compounds being tested in subsequent bioassays to those that were meta- and para- substituted by mid-sized (e.g., butyl-allyl) chains. However, I was not able to detect as much behavioral bioactivity as was perhaps expected based upon previous work conducted with Trichoplusia ni, the cabbage looper (Akhtar et al., 2007, Akhtar et al., 2010). This resulted in the decision to test more species of stored-product beetles for feeding deterrence responses to these compounds. 138 There was a strong correlation between the feeding detterence and the mortality observed in S. oryzae. The feeding bioassays showed that at both doses tested, longer chains (pentyl and butyl) resulted in increases in both feeding deterence and mortality. DEET effectivly deterred feeding but several of the compounds I tested were also effective such as 3c{4,4}, 3c{4,5}, 3c{5,5}, 3c{5,6}, and 3c{N2,O3}. I predicted that using a species more specialized in its host preferences may have resulted in a stronger response to these compounds due to increased sensitivity to deterrent signals or a lower ability to successfully metabolize the compounds (thus resulting in reduced feeding). This has been previously observed with a variety of phytophagous insects and plant secondary metabolites, with generalists generally better able to deal with novel chemistry (Castells and Berenbaum, 2008; Bernays et al., 2000; Haylon et al., 2015). The large number of cytochromes P450s (often associated with detoxification abilities) found in the T. castaneum genome (Tribolium Genome Sequencing Consortium, 2008) lent support to the prediction that T. castaneum (a generalist) is able to successfully metabolize the compounds and continue feeding. This prediction was supported by the mortality results and to a lesser extent the feeding deterrence results from T. castaneum and the closely related species T. confusum (wider range of materials used) compared to Sitophilus orzyae, S. zeamais, and R. dominica (narrower range of materials used). Of the eight compounds tested, S. oryzae, S. zeamais, and R. dominica generally had LT50s that were significantly lower than the control (7, 8, and 7 treatment compounds, respectively) compared to T. castaneum and T. confusum (2 and 0 treatment compounds, respectively). In addition, comparing the food consumed, more treatments reduced feeding by S. oryzae, S. zeamais, and R. dominica to less than 25% of that consumed by controls (6, 4, and 4 treatment compounds, respectively) compared to the Tribolium spp. (1 and 2 treatment 139 compounds). This supports the prediction that Tribolium spp., which uses a wide range of materials, is able to successfully survive on these novel compounds compared to the more Sitophilus spp. and R. dominica which generally feed and reproduce in a narrower range of materials. Further, by testing two groups of very closely related species I was able to highlight the variability in feeding deterrents and in this case also mortality that was observed even between closely related species. For example, T. castaneum showed high and rapid mortality on the flour disks treated with 3c{3,6} but T. confusum did not. Tribolium castaneum feeding was reduced significantly more than feeding by T. confusum on disks treated with 3c{4,4}. There were also significant differences between percent feeding by the Sitophilus spp. at the lower dose tested (e.g., 3c{3,3}, 3c{n5,n5}). Thus caution must be taken when trying to extrapolate bioactivity from one species of insect to another, even a closely related species. It has been noted that one of the downsides of antifeedants as a control tactic is interspecific variability (Isman, 2002), which is supported by the results of these bioassays. This may be a challenge when trying to control stored-product insects as there can be many species all living sympatrically (Arbogast and Throne, 1997; Athanassiou et al., 2005). While I was able to identify feeding deterrent bioactivity for several compounds, I did not determine the mechanism by which feeding was being reduced. Therefore, I examined a subset of the previously tested compounds for repellency. Using a walking bioassay I found that T. castaneum was unaffected by any of the treatments at the dose tested but S. oryzae did show some avoidance of both DEET and 3c{N3,O2}. I was able to detect significant results, even with a low number of replicates due to limited amounts of these synthetic compounds. Both of those 140 compounds also showed feeding deterrence and high mortality to S. orzyae. As those compounds have toxic effects, prefeeding avoidance would increase the insect’s chance of survival if it reduced contact with the compounds, and it seems the beetle may be able to detect the compounds through olfaction. This indicates that some compounds like the ones tested have more than one mode of action with S. oryzae although, with the mortality observed, I cannot rule out post-ingestive toxicity rather than a stimulation of deterrent gustatory receptors (taste) as the mechanism of feeding deterrence. Finally, during the course of preparing bioassays I developed an alternative method of measuring how much of the flour disks were being eaten by the insects (Clark et al., 2014). The method developed by Xie et al. (1996) used weight (before and after) as the measure of insect feeding and I used this method for the bioassays in this thesis. However, that method requires a very sensitive balance, which is not accessible for all researchers due to cost. Therefore, I tested using area of the flour disk fed on as a measure of insect feeding rather than weight. In this method I used a scanner as well as a digital camera and an open-source program to determine the number of image-pixels remaining in the disk to calculate the area fed upon. I found that the gravimetric method (Xie et al., 1996) correlated well with both the scanner and the use of digital photographs. This indicates that these methods are an acceptable and less expensive way of determining insect feeding. This research is a step in developing novel tactics in pest management strategies designed for the protection of commodities against stored-product pests. Any chemical that would be used would need to be registered for commercial application and rigorous testing must occur to determine its suitability, including (but not limited to) vertebrate toxicity, residual activity, 141 potential for water contamination, and where and when it can be best used. I have shown that there is potential for some of these compounds to be used as feeding deterrents that may be effective against several of the most destructive stored-product insects, particularly if used as part of a larger and diverse integrated pest management strategy. 142 Literature Cited Abdelghany, A.Y., Awadalla, S.S., Abdel-Baky, N.F., EL-Syrafi, H.A. and P.G. Fields (2016). Efficacy of reduced risk insecticides on penetration into jute and polyethylene bags by Lasioderma serricorne (F.) (Coleoptera: Anobiidae). Journal of Stored Products Research 69: 190-194. Adarkwah, C., Fields, P.G., Schöeller, M., and C. Reichmuth (2012). Resistance and physiological differences in the rice weevil Sitophilus oryzae (L.) (Coleopteran: Curculionidae) exposed to diatomaceous earth. Mitteilungen der Deutschen Gesellschaft für allgemeine und angewandte Entomologie 18: 433-438. Aebischer, N.J. (1990). Assessing pesticide effects on non-target invertebrates using long-term monitoring and time-series modeling. Functional Ecology 4: 369-373. Ahmed, H. and B.G. Gardiner (1970). Penetration of malathion into locusts. Pesticide Science 1: 217-219. Akhtar, Y. and M.B. Isman (2003). Binary mixtures of feeding deterrents mitigate the decrease in feeding deterrent response to antifeedants following prolonged exposure in the cabbage looper, Trichoplusia ni (Lepidoptera: Noctuidae). Chemoecology 13: 177-182. Akhtar, Y., Rankin, C.H., and M.B. Isman (2003). Decreased response to feeding deterrents following prolonged exposure in the larvae of a generalist herbivore, Trichoplusia ni (Lepidoptera: Noctuidae). Journal of Insect Behavior 16: 811-831. 143 Akhtar, Y., Yeoung, Y.-R., and M.B. Isman (2008). Comparative bioactivity of selected extracts from Meliacea and some commercial botanical insecticides against two noctuid caterpillars, Trichoptera ni and Pseudaletia unipuncta. Phytochemistry Reviews 7: 77-88. Akhtar, Y., Isman, M.B., Paduraru, P.M., Nagabandi, S., Nair, R., and E. Plettner (2007). Screening of dialkoxybenzenes and disubstituted cyclopentene derivatives against the cabbage looper, Trichoplusia ni, for the discovery of new feeding and oviposition deterrents. Journal of Agricultural and Food Chemistry 55: 10323-10330. Akhtar, Y., Yu, Y., Isman, M.B., and E. Plettner (2010). Dialkoxybenzene and dialkoxyallylbenzene feeding and oviposition deterrents against the cabbage looper, Trichoplusia ni: Potential insect behavior control agents. Journal of Agriculture Food Chemistry 58: 49834991. Aliyu, A.B., Adeyemi, M.M.H., Abdulkadir, I., Dambatta, M.B., Amupitan, J.O. and A.O. Oyewale (2014). Antifeedant activity of Vernonia oocephala against stored product pest Tribolium castaneum (Herbst). Bangladesh Journal of Scientific and Industrial Research 49: 243-248. Arbogast, R.T. and Flaherty, B.R. (1973). Light response of Tribolium castaneum and Tribolium confusum (Coleoptera, Tenebrionidae): Variation with age and sex. Journal of Stored Products Research 9: 31-35. Arbogast, R.T. and J.E. Throne (1997). Insect infestation of farm-stored maize in South Carolina: towards characterization of a habitat. Journal of Stored Product Research 33: 187-198. 144 Arthur, F.H. and B. Subramanyam (2012). Chemical control in stored products. In: Hagstrum, D.W., Phillips, T.W., and G. Cuperus (Eds.). Stored Product Protection, Kansas State Research and Extension, Kansas, Chapter 9. Athanassiou, C.G., Kontodimas, D.C., Kavallieratos, N.G., and M.A. Veroniki (2005). Insecticidal effect of NeemAzal against three stored-product beetle species on rye and oats. Journal of Economic Entomology 98: 1733-1738. Athanassiou, C.G., Kavallieratos, N.G., Throne, J.E. and C.T. Nakas (2014). Competition among species of stored-product psocids (Psocoptera) in stored grain. PLoS One 9: e102867. DOI 10.1371/journal.pone.0102867 Bell, W.J. (1990). Searching behavior patterns in insects. Annual Review of Entomology 35: 447467. Bellinger, R.G. (1996). Pest resistance to pesticides. Southern Extension and Research Activity – Information Exchange Group 1. http://ipm.ncsu.edu/safety/factsheets/resistan.pdf (Accessed September 20, 2008). Benton, R., Vannice, K.S., Gomez-Diaz, C., and L.B. Vosshall (2009) Variant ionotropic glutamate receptors as chemosensory receptors in Drosophila. Cell 136: 149-162. doi: 10.1016/j.cell.2008.12.001 Berenbaum, M.R. (1986). Postingestive effects of phytochemicals on insects: On Paracelsus and plant products. In: Miller, T.A. and J. Miller (Eds). Insect-Plant Interactions Springer-Verlag, New York: 159-173. 145 Bernays, E.A., Oppenheim, S., Chapman, R.F., Kwon, H., and F. Gould (2000). Taste sensitivity of insect herbivores to deterrents is greater in specialists than in generalists: a behavioral test of the hypothesis with two closely related caterpillars. Journal of Chemical Ecology 26: 547-563. Bhonwong, A., Stout, M.J., Attajarusit, J., and P. Tantasawat (2009). Defensive role of tomato polyphenol oxidases against cotton bollworm (Helicoverpa armigera) and beet army worm (Spodoptera exigua). Journal of Chemical Ecology 35: 28-38 Bomford, M.K. and M.B. Isman (1996). Desensitization of fifth instar Spodoptera litura to azadirachtin and neem. Entomologia experimentalis et applicata 81: 301-313. Borg-Karlson, A.-K., Nordlander, G., Mudalige, A., Nordenhem, H. and C.R. Unelius (2006). Antifeedants in the feces of the pine weevil Hylobius abietis: Identification and biological activity. Journal of Chemical Ecology 32: 943-957. Bowdan, E. and V.G. Dethier (1986). Coordination of a dual inhibitory system regulating feeding behavior in the blowfly. Journal of Comparative Physiology A 158: 713-722. Braga, L.S., Corrêa, A.S., Pereira, E.J.G., and R.N.C. Guedes (2011). Face or flee? Fenitrothion resistance and behavioral response in populations of the maize weevil, Sitophilus zeamais. Journal of Stored Products Research 47: 161-167. Brower, J.H., Smith, L., Vail, P.V., and P.W. Flinn (1996). Biological control. In: Subramanyam, B. and D.W. Hagstrum (Eds.) Integrated management of insects in stored products. Marcel Dekker, Inc., New York: Chapter 6. Bruce, T.J.A., Wadhams, L.J., and C.M. Woodcock (2005). Insect host location: a volatile situation. Trends in Plant Science 10: 269-274. 146 Burks, C.S. and L.P.S. Kuenen (2012). Effect of mating disruption and lure load on the number of Plodia interpunctella (Hübner) (Lepidoptera: Pyralidae) males captured in pheromone traps. Journal of Stored Products Research 49: 189-195. Burkholder, W.E. and M. Ma (1985). Pheromones for monitoring and control of stored-product insects. Annual Review of Entomology 30: 257-272. Cameron, L.M., Rogers, M., Aalhus, M., Seward, B., Yu, Y., and E. Plettner (2014). Feeding deterrence of cabbage looper (Lepidoptera: Noctuidae) by 1-allyloxy-4-propoxybenzene, alone and blended with neem extract. Journal of Economic Entomology 107: 2119-2129. Campbell, J.F. (2012). Attraction of walking Tribolium castaneum adults to traps. Journal of Stored Products Research 51: 11-22. Campbell, J.F., Mullen, M.A., and A.K. Dowdy (2002). Monitoring stored-product pests in food processing plants with pheromone trapping, contour mapping, and mark-recapture. Journal of Economic Enomology 95: 1089-1101. Campos, M. and T.W. Phillips (2014). Attract-and-kill and other pheromone-based methods to suppress populations of the Indianmeal moth (Lepidoptera: Pyralidae). Journal of Economic Entomology 107: 473-480. Canada Grain Act (1996). Canada Grain Regulations. Canada Gazette, Part II, Vol. 109, No. 14. Cardé, R.T. and A.K. Minks (1995). Control of moth pests by mating disruption: Successes and constraints. Annual Review of Entomology 40: 559-585. 147 Carpinella, M.C., Defago, M.T., Valladares, G., and S.M. Palacios (2003). Antifeedant and insecticide properties of a limonoid from Melia azedarach (Meliaceae) with potential use for pest management. Journal of Agriculture and Food Chemistry 51: 369-374. Castells, E. and M.R. Berenbaum (2008). Resistance of the generalist moth Trichoplusia ni (Noctuidae) to a novel chemical defense in the invasive plant Conium maculatum. Chemoecology 18: 11-18. Cerejeira, M.J., Viana, P., Batista, S., Pereira, T., Silva, E., Valério, M.J., Silva, A., Ferreira, M., and A.M. Silva-Fernandes (2003). Pesticides in Portuguese surface and ground waters. Water Research 37: 1055-1063. Champ, B.R. and C.E. Dyte (1977). FAO Global survey of pesticide susceptibility of stored grain pests. FAO Plant Protection Bulletin 25: 49-67. Chapman, R.F. (2003). Contact chemoreception in feeding by phytophagous insects. Annual Review of Entomology 48: 455-484. Chu, S.S., Wang, C.F., Du, S.S., Lui, S.L., and Z.L. Liu (2011). Toxicity of the essential oil of Illicium difengpi stem bark and its constituent compounds towards two grain storage insects. Journal of Insect Science 11: 1-10. Cis, J., Nowak, G. and W. Kisiel (2006). Antifeedant properties and chemotaxonomic implications of sesquiterpene lactones and syringin from Rhaponticum pulchrum. Biochemical Systematics and Ecology 34: 862-867. Clark, E.L., Isitt, R., Plettner, E., Fields, P.G., and D.P.W. Huber (2014). An inexpensive feeding bioassay technique for stored-product insects. Journal of Economic Entomology 107: 455-461. 148 Collins, P.J. and D. Wilson (1986). Insecticide resistance in the major coleopterous pests of stored grain in southern Queensland. Queensland Journal of Agricultural and Animal Sciences 43: 107-114. Cook, S.M., Khan, Z.R., and J.A. Pickett (2007). The use of push-pull strategies in integrated pest management. Annual Review of Entomology 52: 375-400. Cooper, J. and H. Dobson (2007). The benefits of pesticides to mankind and the environment. Crop Protection 26: 1337-1348. Corrêa, A.S., Pereira, E.J.G., Cordeiro, E.M.G., Braga, L.S., and R.N.C. Guedes (2011). Insecticide resistance, mixture potentiation and fitness in populations of the maize weevil (Sitophilus zeamais). Crop Protection: 1655-1666. Cottee, P.K., Bernays, E.A., and A.J. Mordue (1988). Comparisons of deterrency and toxicity of selected secondary plant compounds to an oligophagous and polyphagous acridid. Entomologia experimentalis et applicata 46: 241-247. Couto, A., Alenius, M., and B.J. Dickson (2005). Molecular, anatomical, and functional organization of the Drosophila olfactory system. Current Biology 15: 1535-1547. Daborn, P.J., Yen, J.L., Bogwitz, M.R., Le Goff, G., Feil, E., Jeffers, S., Tijet, N., Perry, T., Heckel, D., Batterham, P., Feyereisen, R., Wilson, T.G., R.H. French-Constant (2002). A single P450 allele associated with insecticide resistance in Drosophila. Science 297: 2253-2256. Dethier, V.G. (1955). The physiology and histology of the contact chemoreceptors of the blowfly. The Quarterly Review of Biology 30: 348-371. 149 Dethier, V.G. (1976). The hungry fly: a physiological study of the behavior associated with feeding. Cambridge, Massachusetts: Harvard University Press. Dethier, V.G., Browne, B.L., and C.N. Smith (1960). The designation of chemicals in terms of the responses they elicit from insects. Journal of Economic Entomology 53: 134-136. Delorme, R., Fournier, D., Chaufaux, J., Cuany, A., Bride, J.M., Auge, D., and J.B. Berge (1988). Esterase metabolism and reduced penetration are causes of resistance to deltamethrin in Spodoptera exigua HUB (Noctuidea: lepidoptera). Pesticide Biochemistry and Physiology 32: 240-246. Denholm, I. and M.W. Rowland (1992). Tactics for managing pesticide resistance in arthropods: Theory and practice. Annual Review of Entomology 37: 91-112. Devine, G.J. and M.J. Furlong (2007). Insecticide use: Contexts and ecological consequences. Agriculture and Human Values 24: 281-306. Desneux, N., Decourtye, A., and J.-M. Delpuech (2007). The sublethal effects of pesticides on beneficial arthropods. Annual Review of Entomology 52: 81-106. Ditzen, M., Pellegrino, M., and L.B. Vosshall (2008). Insect odorant receptors are molecular targets of the insect repellent DEET. Science 319: 1838-1842. Dong, K. (2007). Insect sodium channels and insecticide resistance. Invertebrate Neuroscience 7: 17-30. Dowdy, A.K., Howard, R.W., Seitz, L.M., and W.H. McGaughey (1997). Response of Rhyzopertha dominica (Coleoptera: Bostrichidae) to its aggregation pheromone and wheat volatiles. Environmental Entomology 22:965-970. 150 Duehl, A.J., Cohnstaedt, L.W., Arbogast, R.T., and P.E.A. Teal (2011). Evaluation light attraction to increase trap efficiency for Tribolium castaneum (Coeloptera: Tenebrionidae). Journal of Economic Entomology 104: 1430-1435. doi: 10.1603/EC10458. Dyte, C.E. and D.G. Blackman (1970). The spread of insecticide resistance in Tribolium castaneum (Herbst) (Coleoptera, Tenebrionidae). Journal of Stored Products Research 6: 255261. Ebrahimi, P., Spooner, J., Weinberg, N., and E. Plettner (2013). Partition, sorption and structure activity relation study of dialkoxybenzenes that modulate insect behavior. Chemosphere 93: 5460. Ecobichon, D.J. (2001). Pesticide use in developing countries. Toxicology 160: 27-33. Edde, P.A. (2012). A review of the biology and control of Rhyzopertha dominica (F.) the lesser grain borer. Journal of Stored Products Research 48: 1-18. El-Sayed, A.M., Suckling, D.M., Byers, J.A., Jang, E.B., and C.H. Wearing (2009). Potential of “lure and kill” in long-term pest management and eradication of invasive species. Journal of Economic Entomology 102: 815-835. Elkinton, J.S., Wood, D.L., and L.E. Browne (1981). Feeding and boring behavior of the bark beetle, Ips paraconfusus, in extracts of ponderosa pine phloem. Journal of Chemical Ecology 7: 209-220. Eriksson, C., Månsson, P.E., Sjödin, K., and F. Schlyter (2008). Antifeedants and feeding stimulants in bark extracts of ten woody non-host species of the pine weevil, Hylobius abietis. Journal of Chemical Ecology 34: 1290-1297. 151 Fairchild, W.L., Swansburg, E.O., Arsenault, J.T., and S.B. Brown (1999). Does an association between pesticide use and subsequent declines in catch of Atlantic salmon (Salmo salar) represent a case of endocrine disruption? Environmental Health Perspectives 107: 349-358. Faustini, D.L., Burkholder, W.E., and R.J. Laub (1981). Sexually dimorphic setiferous sex patch in the male red flour beetle, Tribolium castaneum (Herbst) (Coleoptera: Tenebrionidae): site of aggregation pheromone production. Journal of Chemical Ecology 7: 465-480. Feng, R. and M.B. Isman (1995). Selection for resistance to azadirachtin in the green peach aphid, Myzus persicae. Experientia 51: 831-833. Fields, P.G. (1992). The control of stored-product insects and mites with extreme temperatures. Journal of Stored Products Research 28: 89-118. Fields, P.G. and W.E. Muir (1996). Physical control. In: Subramanyam, B. and D.W. Hagstrum (Eds.) Integrated management of insects in stored products. Marcel Dekker, Inc., New York: Chapter 5. Fields, P.G., Woods, S., and W.G. Taylor (2010) Triterpenoid saponins synergize insecticidal pea peptides: effect on feeding and survival of Sitophilus oryzae (Coleoptera: Curculionidae). The Canadian Entomologist 142: 501-512. Fischer, J., Müller T, Spatz, A-K, Greggers, U, Grünewald, B, and R. Menzel (2014). Neonicotinoids interfere with specific components of navigation in honeybees. PLoS ONE 9: e91364. doi:10.1371/journal.pone.0091364 152 Forstner, M., Breer, H., and J. Krieger (2009). A receptor and binding protein interplay in the detection of a distinct pheromone component in the silkmoth Antheraea polyphemus. International Journal of Biological Sciences 5: 745-757. Foster, S.P. and M.O. Harris (1997) Behavioral manipulation methods for insect pestmanagement. Annual Review of Entomology 42: 123-146. Frazier J.L. and S. Chyb (1995). Use of feeding inhibitors in insect control. In: Chapman, R.F. and G. de Boer (Eds.) Regulatory mechanisms in insect feeding. Chapman and Hall, New York: 364-381. Gabrys, B., Halarewicz-Pacan, A., Nawrot, J., Pradzynska, A., Aniol, M., Szumny, A., and C. Wawrzenczyk (2001). Chromenes, chromanones and alkyl substituted phenols as antifeedants to storage pests and aphids. Journal of Plant Protection Research 41: 229-239. Georghiou, G.P. (1972). The evolution of resistance to pesticides. Annual Review of Ecology, Evolution, and Systematics 3: 133-168. Germinara, G.S., Rotundo, G., and A. De Cristofaro (2007). Repellence and fumigant toxicity of propionic acid against adults of Sitophilus granarius (L.) and S. oryzae (L.). Journal of Stored Product Research 43: 229-233. Glendinning, J.I. (1996). Is chemosensory input essential for the rapid rejection of toxic foods? The Journal of Experimental Biology 199: 1523-1534. Glendinning, J.I. and F. Slansky Jr. (1994). Interactions of allelochemicals with dietary constituents: effects on deterrency. Physiological Entomology 19: 173-186. 153 GNU Image Manipulation Program (2000). GNU Image Manipulation Program v. 2.6.11 for Windows. www.gimp.org Gould, F. (1984). Role of behavior in the evolution of insect adaptation to insecticides and resistant host plants. Bulletin of the Entomological Society of America 30: 34-41. Guedes, R.N.C., Lima, J.O.G., Santos, J.P., and C.D. Cruz (1995). Resistance to DDT and pyrethroids in Brazilian populations of Sitophilus zeamais Motsch. (Coleoptera: Curculionidae). Journal of Stored Products Research 31: 145-150. Guedes, N.M.P., Guedes, R.N.C., Ferreira, G.H. and L.B. Silva (2009). Flight take-off and walking behavior of insecticide-susceptible and –resistant strains of Sitophilus zeamais exposed to deltamethrin. Bulletin of Entomological Research 99: 393-400. Guedes, R.N.C, Oliveira, E.E., Guedes, N.M.P., Ribeiro, B., and J.E. Serrão (2006). Cost and mitigation of insecticide resistance in the maize weevil, Sitophilus zeamais. Physiological Entomology 31: 30-38. Guo, M., Krieger, J., Grobe-Wilde, E., Mibbach, C., Zhang, L., and H. Breer (2014). Variant ionotropic receptors are expressed in olfactory sensory neurons of coeloconic sensilla on the antenna of the desert locusts (Schistocerca gregaria). International Journal of Biological Sciences 10: 1-14. doi: 10.7150/ijbs.7624 Hagstrum D.W. and E.E. Gilbert (1976). Emigration rate and age structure of Tribolium castaneum populations during growth phase of a colonizing episode. Environmental Entomology 5: 445-448. 154 Hagstrum, D.W. and B. Subramanyam (2006). Fundamentals of stored-product entomology. AACC International: Chapter 6 & 7. Hagstrum, D.W. and B. Subramanyam (2012). Monitoring and decision tools. In: Subramanyam, B. and D.W. Hagstrum (Eds.) Alternatives to Pesticides in Stored-Product IPM, Springer Science & Business Media: Chapter 1. Hagstrum, D.W. Reed, C., and P. Kenkel (1999). Management of stored wheat insect pests in the USA. Integrated Pest Management Reviews 4: 127-142. Hahn, J., Jesse, L., and P. Pellitteri (accessed 2016). Insect pests of stored foods. University of Minnesota Extension. http://www.extension.umn.edu/garden/insects/find/insect-pests-of-storedfood/ Hansen, E., and M. Donohoe (2003). Health issues of migrant and seasonal farmworkers. Journal of Health Care for the Poor and Underserved 14: 153-164. Haynes, K.F. (1988). Sublethal effects of neurotoxic insecticides on insect behavior. Annual Review of Entomology 33:149-168. Highland, H.A. (1984). Insect infestation of packages. In: Baur, F.J. (Ed.). Insect Management for Food Processing. American Association of Cereal Chemists, St. Paul, MN: 311-320. Ho, S.H., Ma, Y., Y. Huang (1997). Anethole, a potential insecticide from Illicium verum Hook F., against two stored product insects. International Pest Control 39: 50-51. Hodges, R.J., Robinson, R., and D.R. Hall (1996). Quinone contamination of dehusked rice by Tribolium castaneum (Herbst) (Coleoptera:Tenebrionidae). Journal of Stored Product Research 32: 31-37. 155 Hokkanen, H.M. (1991). Trap cropping in pest management. Annual Review of Entomology 36: 119-138. Honson, N., Johnson, M.A., Oliver, J.E., Prestwich, G.D., and E. Plettner (2003). Structureactivity studies with pheromone-binding proteins of the gypsy moth, Lymantria dispar. Chemical Senses 28: 479-489. Hou, X. and P. Fields (2003). Effectiveness of protein-rich pea flour for the control of storedproduct beetles. Entomologia Experimentalis et Applicata 108: 125-131. Hou, X., Fields, P., and W. Taylor (2004). Combination of protein-rich pea flour and pea extract with insecticides and enzyme inhibitors for control of stored-product beetles. The Canadian Entomologist 136: 581-590. Hou, X., Fields, P., and W. Taylor (2004). The effect of repellents on penetration into packaging by stored-product insects. Journal of Stored Products Research 40: 47-54. Howe, R.W. (1956). The effect of temperature and humidity on the rate of development and mortality of Tribolium castaneum (Herbst) (Coleoptera, Tenebrionidae). Ann. appl. Biol. 44: 356-368. Huang, Y., Tan, J.M.W.L., Kini, R.M. and S.H. Ho (1997). Toxic and antifeedant action of nutmeg oil against Tribolium castaneum (Herbst) and Sitophilus zeamais Motsch. Journal of Stored Product Research 33: 289-298. Huang, Y. and S.H. Ho (1998). Toxicity and antifeedant activities of cinnamaldehyde against the grain storage insect, Tribolium castaneum (Herbst) and Sitophilus zeamais Montsch. Journal of Stored Product Research 34: 11-17. 156 Huang, Y., Ho, S.H., and R.M. Kini (1999). Bioactivities of safrole and isosafrole on Sitophilus zeamais (Coleoptera: Curculionidae) and Tribolium castaneum (Coleoptera: Tenebrionidae). Journal of Economic Entomology 92: 676-683. Huang, Y., Lam, S.L., and S.H. Ho (2000). Bioactivities of essential oil from Elletara cardamomum (L.) Maton. to Sitophilus zeamais Motschulsky and Tribolium castaneum (Herbst). Journal of Stored Product Research 36: 107-117. Huang, Y., Ho, S.-H. Lee, H.-C., and Y.-L. Yap (2002). Insecticidal properties of eugenol, isoeugenol and ethyleugenol and their effects on nutrition of Sitophilus zeamais Motsch. (Coleoptera: Curculionidae) and Tribolium castaneum (Herbst) (Coleoptera: Tenebrionidae). Journal of Stored Products Research 38: 403-412. Hunt, T., Bergsten, J., Levkanicova, Z., Papadopoulou, A., St.John, O., Wild, R., Hammond, P.M., Ahrens, D., Balke, M., Caterion, M.S., Gómez-Zurita, J., Ribera, I., Barraclough, T.G., Bocakova, M., Bocak, L., and A.P. Vogler (2007). A comprehensive phylogeny of beetles reveals the evolutionary origins of a superradiation. Science 318: 1913-1916. Ishak, I.H., Riveron, J.M., Ibrahim, S.S., Stott, R., Longbottom, J., Irving, H., and C.S. Wondji (2016). The cytochrome P450 gene CYP6P12 confers pyrethroid resistance in kdr-free Malaysian populations of the dengue vector Aedes albopictus. Scientific Reports April. DOI: 10.1038/srep24707. Isman, M.B. (1993). Growth inhibitory and antifeedant effects of azadirachtin on six noctuids of regional economic importance. Pesticide Science 38: 57-63 157 Isman, M.B. (2002). Insect antifeedants. Pesticide Outlook Aug: 152-157. DOI: 10.1039/b206507j. Isman, M.B. (2006) Botanical insecticides, deterrents, and repellents in modern agriculture and an increasing regulated world. Annual Review of Entomology 51: 45-66. Isman, M.B.,Matsuura, H., MacKinnon, S., Durst, T., Towers, G. H. N., and J.T. Arnason (1996). Phytochemistry of the Meliacea: So many terpenoids, so few insecticides. Recent Advances in Phytochemistry 30: 155-178. Jeon, J.-H., Oh, M.-S., Cho, K.-S. and H.-S. Lee (2012). Phototactic response of the rice weevil, Sitophilus oryzae Linnaeus (Coleoptera: Curculionidae), to light-emitting diodes. Journal of the Phototactic response of the rice weevil, Sitophilus oryzae Linnaeus (Coleoptera: Curculionidae), Korean Society for Applied Biological Chemistry 55: 35-39. Khan, M.A. and R. Wohlgemuth (1980). Diethyltoluamide as a repellent to stored product insects. Anzeiger fur Schadlingskunde Pflanzenschutx Umweltschutz 53: 126-127. Koul, O. (2004). Insect Antifeedants. CRC Press, Boca Raton, FL. Koul, O. (2008). Phytochemicals and insect control: an antifeedant approach. Critical Reviews in Plant Sciences 27: 1-24. Koul, O., Multani, J.S., Singh, G., Daniewski, W.M. and S. Berlozecki (2003). 6betahydroxygedunin from Azadirachta indica. Its potentiation effects with some non-azadirachtin limonoids in neem against lepidopteran larvae. Journal of Agriculture and Food Chemistry 7: 2937-2942. 158 Koul, O., Singh, G., Singh, R., Singh, J., Daniewski, W.M., and S. Berlozecki (2004). Bioefficacy and mode-of-action of some limonoids of salannin group from Azadirachta indica A. Juss and their role in a multicomponent system against lepidopteran larvae. Journal of Bioscience 29: 409-416. Larsson, M.C., Domingos, A.I., Jones, W.D., Chiappe, M.E., Amrein, H., and L.B. Vosshall (2004). Or83b encodes a broadly expressed odorant receptor essential for Drosophila olfaction. Neuron 43: 703-714. Lattanzio, V., Lattanzio, V.M.T., and A. Cardinali (2006). Role of phenolics in the resistance mechanisms of plants against fungal pathogens and insects. Phytochemistry: Advances in Research: 23-67. Lee, S.E. (2002). Biochemical mechanisms conferring cross-resistance to fumigant toxicities of essential oils in a chlorpyrifos-methyl resistant strain of Oryzaephilus surinamensis L. (Coleoptera: Silvanidae). Journal of Stored Products Research 38: 157-166. Leeper, J.R., Roush, R.T., and H.T. Reynolds (1986). Preventing or managing resistance in arthropods. In: Committee on strategies for the management of pesticide resistant pest population (Eds.). Pesticide Resistance: Strategies and tactics for managements. National Academy Press, Washington, D.C.: 335-346. Levin, D.A. (1971). Plant phenolics: an ecological perspective. The American Naturalist 105: 157-181. Levison, H.Z. and K. Mori (1983) Chirality determines pheromone activity for flour beetles. Naturwissenschaften 70: 190-192. 159 Levinson, H.Z., Levinson, A., Ren, Z., K. Mori (1990). Comparative olfactory perception of the aggregation pheromones of Sitophilus oryzae (L.), S. zeamais (Motsch.) and S. granaries (L.), as well as the stereoisomers of these pheromones. Journal of Applied Entomology: 110: 203-213. Likhayo, P.W. and R.J. Hodges (2000). Field monitoring Sitophilus zeamais and Sitophilus oryzae (Coleoptera: Curculionida) using refuge and flight traps baited with synthetic phermone and cracked wheat. Journal of Stored Products Research 36: 341-353. Lindgren, D.L., Vincent, L.E., Krohne, H.E. (1954). Relative effectiveness of ten fumigants to adults of eight species of stored-product insects. Journal of Economic Entomology 47: 923-926. Lindgren, B.S., Borden, J.H., Pierce, A.M., Pierce, H.D., Oehlschlager, A.C., and J.W. Wong (1985). A potential method for simultaneous, semiochemical-based monitoring of Cryptolestes ferrugineus and Tribolium castaneum (Coleoptera: Cucujidae and Tenebrionidae). Journal of Stored Products Research 21: 83-87. Lindroth, R.L. (1991). Differential toxicity of plant allelochemicals to insects: roles of enzymic detoxication system. In: E.A. Bernays (Ed.). Insect-plant interactions, Vol III. CRC Press, Boca Raton, Florida: 1-34. Linsley, E.G. (1944). Natural sources, habitats, and reservoirs of insects associated with stored food products. Hilgardia 16: 187-224. Liu, N. and J.G. Scott (1998). Increased transcription of CYP6D1 causes cytochrome P450mediated insecticide resistance in house fly. Insect Biochemistry and Molecular Biology 28: 531535. 160 Liu, Z.L. and S.H. Ho. (1999). Bioactivity of the essential oil extracted from Evodia rutaecarpa Hook f. et Thomas against the grain storage insects, Sitophilus zeamais Motsch. and Tribolium castaneum (Herbst). Journal of Stored Products Research 35: 317-328. Markarian, H., Florentine, G. J., and J.J. Pratt, Jr. (1978). Quinone production of some species of Tribolium. Journal of Insect Physiology 24: 785-790. Mason, L.J. (2003). Red and confused flour beetles Tribolium castaneum (Bhst.) and Tribolium confusm Duval. Grain Insect Fact Sheet E-224-W, Purdue University. Mason, L.J. and M. McDonough (2012). Biology, behavior, and ecology of stored grain and legume insects. In: Hagstrum, D.W., Phillips, T.W., and G. Cuperus (Eds.) Stored Product Protection. Kansas State Research and Extension, Kansas: 7-20. Matlock, Jr., R.B. and R. de la Cruz (2002). An inventory of parasitic Hymenoptera in banana plantations under two pesticide regimes. Agriculture, Ecosystems and Environment 93: 147-164. Maxwell, C.A., Msuya, E., Sudi, M., Njunwa, K.J., Carneiro, I.A., and C.F. Curtis (2002). Effect of community-wide use of insecticide-treated nets for 3-4 years on malarial morbidity in Tanzania. Tropical Medicine and International Health 7: 1003-1008. Miller, J.R., Siegert, P.Y., Amimo, F.A., and E.D. Walker (2009). Designation of chemicals in terms of the locomotor responses they elicit from insects: An update of Dethier et al., (1960). Journal of Economic Entomology 102: 2056-2060. Mitchell, E.R., Hu, G., and D. Johanowicz (2000). Management of diamondback moth (Lepidoptera: Plutellidae) in cabbage using collard as a trap crop. HortScience 35: 875-879. 161 Mochizuki, M. (2003). Effectiveness and pesticide susceptibility of the pyrethroid-resistant predatory mite Amblyseius womersleyi in the integrated pest management of tea pests. BioControl 48: 207-221. Mohanty, S., Zubkov, S., and A.M. Gronenborn (2004). The solution NMR structure of Antheraea Polyphemus PBP provides new insight into pheromone recognition by pheromonebinding proteins. Journal of Molecular Biology 337: 443-451. Mordue (Luntz), A.J. (2007). Present conecepts of the mode of action of azadirachtin from neem. In: Koul, O. And S. Wahab (Eds). Neem: Today and in the New Millennium, Spinger Science & Business Media: Chapter 11. Mordue, W., Goldsworthy, G.J., Brady, J., and W.M. Blaney (1980). Insect Physiology. Blackwell Scientific Publications, Boston. Morgan, E.D. (2009). Azadirachtin, a scientific gold mine. Bioorganic and Medicinal Chemistry 17: 4096-4105. Mori, K., Nagao, H., and Y. Yoshihara (1999). The olfactory bulb: Coding and processing of odor molecule information. Science 286: 711-715. Morrissey, C.A., Albert, C.A., Dods, P.L., Cullen, W.R., Lai, V.W.-M. and J.E. Elliott (2007). Arsenic accumulation in bark beetles and forest birds occupying mountain pine beetle infested stands treated with monosodium methanearsonate. Environmental Science and Technology 41: 1491-1500. 162 Mota-Sanchez, D., Hollingworth, R.M., Grafius, E.J., and D.D. Moyer (2006). Resistance and cross-resistance to neonicotinoid insecticides and spinosad in the Colorado potato beetle, Leptinotarsa decemlineata (Say) (Coleoptera: Chrysomelidae). Pest Management Science 62: 30-37. Mullen, M.A. (1994). Development of pheromone-baited insect traps. In: Highley, E., wright, E.J., Banks, H.J., Champ, B.R. (Eds.) Proceedings of the 6th International Working Conference on Stored-product Protection Canberra, Australia. CAB International: 421-424. Mullen, M.A., Vardeman, J.M., and J. Bagwell (2012). Insect resistant packaging. In: Hagstrum, D.W., Phillips, T.W. and G. Cuperus (Eds.). Stored Product Protection. Kansas State Research and Extension, Kansas: Chapter 12. Müller, P., Warr, E., Stevenson, B.J., Pignatelli, P.M., Morgan, J.C., Steven, A., Yawson, A.E., Mitchell, S.N., Ranson, H., Hemingway, J., Paine, M.J.I., and M. J. Donnelly (2008). Fieldcaught permethrin-resistant Anopheles gambiae overexpress CYP6P3, a P450 that metabolises pyrethroids. PLoS Genetics 4: e1000286. doi:10.1371/journal.pgen.1000286 Murlis, J., Elkinton, J.S., and R.T. Cardé (1992). Odor plumes and how insects use them. Annual Review of Entomology 37: 505-532. Myers, J.H., Simberloff, D., Kuris, A.M., and J.R. Carey (2000). Eradication revisited: dealing with exotic species. Trends in Ecology and Evolution 15: 316-320. Nawrot, J., Harmatha, J., and E. Bloszyk (1986). Secondary plant metabolites with antifeeding activity and their effect on some stored product insects. In: 4th International Work Conference on Stored-Product Protection. Tel Aviv, Israel: 591-597. 163 Naylor, A.F. (1961). Dispersal in the red flour beetle Tribolium castaneum (Tenebrionidae). Ecology 42: 231-237. Nicholson, R.L. and R. Hammerschmidt (1992). Phenolic compounds and their role in disease resistance. Annual Review of Phytopathology 30: 369-389. Noack, S. and H.U. Schmidt (1981). Studies on the effect of repellents and insecticides used for the impregnation of packaging materials for protection against stored-product pests. Zeitschriftfur-Angewandte-Entomologie 922: 202-212. Nyamador, W.S., Ketoh, G.K., Amévoin, K., Nuto, Y., Koumaglo, H.K., and I.A. Glitho (2009). Variation in the susceptibility of two Callosobruchus species to essential oils. Journal of Stored Products Research, DOI: 10.1016/j.jspr.2009.09.002. O’Neal, M.E., Landis, D.A., and R. Isaacs (2002). An inexpensive, accurate method for measuring leaf area and defoliation through digital image analysis. Journal of Economic Entomology 95: 1190-1194. Obeng-Ofori, D. and C.H. Reichmuth (1997). Bioactivity of eugenol, a major component of essential oil of Ocimum suave (Wild.) against stored-product Coleoptera. International Journal of Pest Management 43: 89-94. Omar, S., Marcotte, M., Fields, P., Sanchez, P.E., Poveda, L., Mata, R., Jimenez, A., Durst, T., Zhang, J., MacKinnon, S., Leaman, D., Arnason, J.T., and B.J.R. Philogène (2007). Antifeedant activities of terpenoids isolated from tropical Rutales. Journal of Stored Products Research 43: 92-96. 164 Onstad, D.W. (2008). The role of environment in insect resistance management. In: Onstead, D.W. (Ed). Insect resistance: Biology, Economics and Prediction. Elsevier, London: 209-226. Opit, G., Collins, P.J. and G.J. Daglish (2012). Chemical control in stored products, In. Hagstrum, D.W., Phillips, T.W., and G. Cuperus (Eds.) Stored Product Protection. Kansas State Research and Extension, Kansas: 143-155. Paduraru, P.M., Popoff, R.T.W., Nair, R., Gries, R., Greis, G., and E. Plettner (2008). Synthesis of substituted alkoxy benzene minilibraries, for the discovery of new insect olfaction or gustation inhibitors. Journal of Combinatorial Chemistry 10: 123-134. Paré, P.W. and J.H. Tumlinson (1999). Plant volatiles as a defense against insect herbivores. Plant physiology 121: 325-331. Park, I.-K., Lee, S.-G., Choi, D.-H., Park, J.-D., and T.-J. Ahn (2003). Insecticidal activities of constituents identified in the essential oil from leaves of Chamaecyparis obtuse against Callosobruchus chinensis (L.) and Sitophilus oryzae (L.). Journal of Stored Products Research 39: 375-384. Perez-Mendoza, J. (1999). Survey of insecticide resistance in Mexican populations of maize weevil Sitophilus zeamais Motschulsky (Coleoptera: Curculionidae). Journal of Stored Products Research 35: 107-115. Perrucci, S., Macchioni, G., Cioni, P.L., Flamini, G., and I. Morelli (1995). Structure/activity relationship of some natural monoterpenes as acaricides against Psoroptes cuniculi. Journal of Natural Products 58: 1261-1264. 165 Phillips, T.W. (1997). Semiochemicals of stored-product insects: Research and applications. Journal of Stored Products Research 33: 17-30. Phillips, T., Thoms, E.M., DeMark, J., and S. Walse (2012). Fumigation. In. D. W. Hagstrum, T. W. Phillips and G. Cuperus (eds.), Stored Product Protection. Kansas State Research and Extension, Kansas: 157-178. Phillips, J.K., Walgenbach, C.A., Klein, J.A., Burkholder, W.E., Schmuff, N.R. and H.M. Fales (1985). (R*,S*)-5-hydroxy-4-methyl-3heptanone male-produced aggregation pheromone of Sitophilus oryzae (L.) and S. zeamais Motsch. Journal of Chemical Ecology 11: 1263-1274. Phillips, T.W., Jiang, X.-L., Burkholder, W.E., Phillips, J.K., and H.Q. Tran (1993). Behavioral responses to food volatiles by two species of stored-product Coleoptera, Sitophilus oryzae (Curculionidae) and Tribolium castaneum (Tenebrionidae). Journal of Chemical Ecology 19: 723-734. Phillips, T.W., Cogan, P.M. and H.Y. Fadamiro (2000). Pheromones. In: Subramanyam, B. and D.W. Hagstrum (Eds.) Alternatives to pesticides in stored-product IPM. Kluwer Academic Publishers, Boston, Massachussetts, USA: 273-302. Pimentel, D. (1995). Amounts of pesticides reaching target pests: environmental impacts and ethics. Journal of Agricultural and Environmental Ethics 8: 17-29. Pimentel, D. (2005). Environmental and economic costs of the application of pesticides primarily in the United States. Environment, Development and Sustainability 7: 229-252. 166 Pimentel, D. Zuniga, R., and D. Morrison (2005). Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics 52: 273288. Pimentel, D., Acquay, H., Biltonen, M., Rice, P., Silva, M., Nelson, J., Lipner, V., Giordano, S., Horowitz, A., and M. D’Amore (1992). Environmental and economic costs of pesticide use. BioScience 42: 750-760. Pimentel, M.A.G., Faroni, L.R.D’A., Guedes, R.N.C., Sousa, A.H., and M.R. Tótola (2009). Phosphine resistance in Brazilian populations of Sitophilus zeamais Motschulsky (Coleoptera: Curculionidae). Journal of Stored Products Research 45: 71-74. Plapp, Jr., F.W. and R.F. Hoyer (1968). Insecticide resistance in the house fly: decreased rate of absorption as the mechanism of action of a gene that acts as an intensifier of resistance. Journal of Economic Entomology 61: 1298-1303. Plettner, E. and R. Gries (2010). Agonists and antagonists of antennal responses of gypsy moth (Lymantria dispar) to the pheromone (+)-disparlure and other odorants. Journal of Agriculture and Food Chemistry 58: 3708-3719. Prestwich, G.D., Du, G. and S. LaForest (1995). How is pheromone specificity encoded in proteins? Chemical Senses 20:461-469. Priesner, E., Jacobson, M., and H.J. Bestmann (1975). Structure-response relationship in noctuid sex pheromone reception. An introductory report. Z. Naturforsch. C. 30: 283-293. Pyke, B., Rice, M., Sabine, B., and M. Zalucki (1987). The push-pull strategy. The Australian Grower May-July: 7-9. 167 R Core Team (2012). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0,. URL http://www.R-project.org. Raffa, K.F. and J.L. Frazier (1988). A generalized model for quantifying behavioral desensitization to antifeedants. Entomologica experimentalis et applicata 46: 93-100. Rajendran, S. (2002). Postharvest pest losses. In: Pimentel, D. (Ed), Encyclopedia of Pest Management, Marcel Dekker, Inc., New York, pp. 654-656. Rees, D.P. (1996). Coleoptera In: B. Subramanyam and D.W. Hagstrum (eds.) Integrated management of insects in stored products, pp. 287-330 New York: Marcel Dekker, Inc. Rees, D. (2004). Insects of stored products. CSIRO Publishing, Collingwood, Australia. Riffell, J.A., Abrell, L., and J.G. Hildebrand (2008). Physical processes and real-time chemical measurement of the insect olfactory environment. Journal of Chemical Ecology 34: 837-853. Ritter, W.F. (1990). Pesticide contamination of ground water in the United States – a review. Journal of Environmental Science and Health, Part B: Pesticides, Food Contaminates, and Agricultural Wastes 25: 1-29. Ryan, M.F. and D.P. O’Ceallachain (1976). Aggregation and sex pheromones in the beetle Tribolium confusum. Journal of Insect Physiology 22: 1501-1503. Rytz, R., Croset, V., and R. Benton (2013). Ionotropic receptors (IRs): Chemosensory ionotropic glutamate receptors in Drosophila and beyond. Insect Biochemistry and Molecular Biology: 110. http://dx.doi.org/10.1016/j.ibmb.2013.02.007 168 Sambaraju, K.R. and T.W. Phillips (2008). Effects of physical and chemical factors on oviposition by Plodia interpunctella (Lepidoptera: Pyralidae). Annals of the Entomological Society of America 101: 955-963. Sawicki, R.M. and K.A. Lord (1970). Some properties of a mechanism delaying penetration of insecticides into houseflies. Pesticide Science 1: 213-217. Schöller, M. (2010). Prospects for biological control of stored-product pests. In. Reichmuth, C. and M. Schöller (Eds.). International European Symposium on Stored Product Protection “Stress on chemical products”. Berlin, Germany 25-26 May, 2009: 25-31. Schöller, M., Flinn, P.W., Grieshop, M.J., and E. Zd’arkova (2006). Biological control of storedproduct pests. In: Heaps, J. (Ed.) Insect Management for Food Storage and Processing, 2nd ed. American Association of Cereal Chemists: 67-87. Semeao, A.A., Campbell, J.F., Whitworth, R.J., and P.E. Sloderbeck (2011). Response of Tribolium castaneum and Tribolium confusum adults to vertical black shapes and its potential to improve trap capture. Journal of Stored Products Research 47: 88-94. Sfara, V., Mougabure-Cueto, G., Zerba, E.N., and R.A. Alzogaray (2011). Adaptation of the repellency response to DEET in Rhodnius prolixus. Journal of Insect Physiology 57:1431-1436. Shaaya, E., Kostjukovski, M., Eilberg, J., and C. Sukprakarn (1997). Plant oils as fumigants and contact insecticides for the control of stored-product insects. Journal of Stored Products Research 33: 7-15. 169 Shields, V.D.C., Smith, K.P., Arnold, N.S., Gordon, I.M., Shaw, T.E., and D. Waranch (2008). The effect of varying alkaloid concentrations on the feeding behavior of gypsy moth larvae, Lymantria dispar (L.) (Lepidoptera: Lymantriidae). Arthropod Plant Interactions 2: 101-107. Shelton, A.M. and F.R. Badenes-Perez (2006). Concepts and applications of trap cropping in pest management. Annual Review of Entomology 51: 285-308. Simmonds, M.S.J., Blaney, W.M., Ley, S.V., Anderson, J.C. and P.L. Toogood (1990). Azadirachtin: structural requirements for reducing growth and increasing mortality in lepidopterous larvae. Entomologia Experimentalis et Applicata 55: 169-181. Sinha, R.N. and F.L. Watters (1985). Insect pests of flour mills, grain elevators, and feed mills and their control. Agriculture Canada, Ottawa: Chapter 11. Snyder, M.J. and J.I. Glendinning (1996). Causal connection between detoxification enzyme activity and consumption of a toxic plant compound. Journal of Comparitive Physiology A. 179: 255-261. Stapel, J.O., Cortesero, A.M., and W.J. Lewis (2000). Disruptive sublethal effects of insecticides on biological control: Altered foraging ability and life span of a parasitoid after feeding on extrafloral nectar of cotton treated with systemic insecticides. Biological Control 17: 243-249. Steinbrecht, R.A. (1997). Pore structures in insect olfactory sensilla: a review of data and concepts. International Journal of Insect Morphology and Embryology 26: 229-245. Stejskal, V. (1995). The influence of food and shelter on the efficacy of a commercial sticky trap in Tribolium castaneum (Coleoptera: Tenebrionidae). Journal of Stored Products Research 31: 229-233. 170 Stengl, M. and N.A. Funk (2013). The role of the coreceptor Orco in insect olfactory transduction. Journal of Comparative Physiology. A, Neuroethology Sensory Neural and Behavioral Physiology 199: 897-909. doi: 10.1007/s00359-013-0837-3 Strong, R.G. and D.E. Sbur (1961). Evaluation of insecticides as protectants against pests of stored grain and seeds. Journal of Economic Entomology 54: 235-238. Su, H.C.F., Horvat, R., and G. Jilani (1982). Isolation, purification, and characterization of insect repellents from Curcuma longa L. Journal of Agriculture and Food Chemistry 30: 290-292. Subramanyam, B. and D. Hagstrum (1995). Resistance measurement and management. In: Subramanyam, B. and D. Hagstrum (Eds.). Integrated Management of Insects in Stored Products. Marcel Dekker Inc., New York: 331-98. Subramanyam, B. and D. Hagstrum (2000). Alternatives to pesticides in stored-product IPM. Kluwer Academic Publishers, Boston, Mass. Suthisut, D., Fields, P.G., and A. Chandrapatya (2011). Contact toxicity, feeding reduction, and repellency of essential oils from three plants from the ginger family (Zingiberaceae) and their major compondents against Sitophilus zeamais and Tribolium castaneum. Journal of Economic Entomology 104: 1445-1454. Suzuki, T. (1981). Identification of the aggregation pheromone of flour beetles Tribolium castaneum and T. confusum (Coleoptera: Tenebrionidae). Agricultural and Biological Chemistry 45: 1357-1363. Swale, D.R., Sun, B., Tong, F., and J.R. Bloomquist (2014). Neurotoxicity and mode of action of N, N-diethyl-meta-toluamide (DEET). PLoS One 9: e103713. 171 Thomson, D.R., Gut, L.J., and J.W. Jenkins (1999). Pheromones for insect control. In: Hall, F.R. and J.J. Menn (Eds.). Biopesticides: Use and delivery. Springer:385-412. DOI: 10.1385/0-89603515-8:385 Trematerra, P. And P. Girgenti (1989). Influence of pheromone and food attractants on trapping of Sitophilus oryzae (L.) (Col., Curcuilionidae): a new trap. Journal of Applied Entomology 108: 12-20. Trematerra, P., Sciarreta, A., and E. Tamasi (2000). Behavioural responses of Oryzaephilus surinamensis, Tribolium castaneum, and Tribolium confusum to naturally and artificially damaged durum wheat kernels. Entomologia Experimentalis et Applicata 94: 195-200. Tribolium Genome Sequencing Consortium (2008). The genome of the model beetle and pest Tribolium castaneum. Nature 452: 949-955. Turner, S.L. and A. Ray (2009). Modification of CO2 avoidance behavior in Drosophila by inhibitory odorants. Nature 461: 277-282. Ukeh, D.A., Birkett, M.A., Bruce, T.J.A., Allan, E.J., Pickett, J.A., A. J. Mordue (2009). Behavioural responses of the maize weevil, Sitophilus zeamais, to host (stored-grain) and nonhost plant volatiles. Pest Management Science 66: 44-50. Van den Berg, H., Hassan, K. and M. Marzuki (1998). Evaluation of pesticide effects on arthropod predator populations in soya bean in farmer’s fields. Biocontrol Science and Technology 8: 125-137. Via, S. (1999). Cannibalism facilitates the use of a novel environment in the flour beetle, Tribolium castaneum. Heredity 82: 267-275. 172 Wanner, K.W., Nichols, A.S., Allen, J.E., Bunger, P.L., Garczynski, S.F., Linn, Jr., C.E., Robertson, H.M., and C.W. Luetje (2010). Sex pheromone receptor specificity in the European corn borer moth, Ostrinia nubilalis. PLoS ONE 5: e8685. DOI: 10.1371/journal.pone.0008685. Watson E. and G. Barson (1996). A laboratory assessment of the behavioral response of three strains of Oryzaephilus surinamensis (L.) (Coleoptera: Silvanidae) to three insecticides and the insect repellent N,N-diethyl-m-toluamide. Journal of Stored Products Research 32: 59-67. Watson, E., Barson, G. Pinniger, D.B., Roberts, G., and A.R. Ludlow (1997) Evaluation of the behavioural responses of Anthrenus verbasci adults and larvae to permethrin (ec) using a computerized tracking system. Journal of Stored Product Research 33: 335-346. White, N.D.G. and J.G. Leesch (1996). Chemical control In: B. Subramanyam and D.W. Hagstrum (eds.) Integrated management of insects in stored products, pp. 1-40 New York: Marcel Dekker, Inc. Wijkamp, I. and D. Peters (1993). Determination of the median latent period of two tospoviruses in Frankliniella occidentalis, using a novel leaf disk assay. Phytophathology. 83: 986-991. Wilson, E.O. (1965). Chemical communication in the social insects. Science 149: 1064-1071. Wondji, C.S., Irving, H., Morgan, J., Lobo, N.F., Collins, F.H., Hunt, R.H., Coetzee, M., Hemingway, J., and H. Ranson (2009). Two duplicated P450 genes are associated with pyrethroid resistance in Anopheles funestus, a major malaria vector. Genome Research 19: 452459. 173 Xie, Y.S., Fields, P.G., and M.B. Isman (1995). Repellency and toxicity of azadirachtin and neem concentrates to three stored-product beetles. Journal of Economic Entomology 88: 10241031. Xie, Y.S., Bodnaryk, R.P. and P.G. Fields (1996). A rapid and simple flour-disk bioassay for testing substances active against stored-product insects. The Canadian Entomologist 128: 865875. Zettler, J.L. (1991). Pesticide resistance in Tribolium castaneum and T. confusum (Coleoptera: Tenebrionidae) from flour mills in the United States. Journal of Economic Entotmology 84: 763767. Zhang, Q.-H. and F. Schlyter (2003). Redundancy, synergism, and active inhibitory range of non-host volatiles in reducing pheromone attraction in European spruce bark beetle Ips typographus. Oikos 101: 299-310. Zhu, F., Parthasarathy, R., Bai, H., Woithe, K., Kaussmann, M., Nauen, R., Harrison, D.A., and S.R. Palli (2010). A brain-specific cytochrome P450 responsible for the majority of deltamethrin resistance in the QTC279 strain of Tribolium castaneum. PNAS 107: 8557-8562. 174 Appendix Appendix A. Description of how test compounds were made in Dr. Plettner’s laboratory at Simon Fraser University. Briefly, test compounds were derived from catechol (1,2-dihydroxybenzene) for the "a" series, resorcinol (1,3-dihydroxybenzene) for the "b" series, or dihydroquinone (1,4dihydroxybenzene) for the "c" series. Dialkoxybenzenes were prepared by alkylation of phenolic groups with primary alkyl halides, as described in Paduraru et al. (2008). The series of compounds labelled 5a, 5b, or 5c was derived from the corresponding allyloxy-alkoxybenzene by Claisen rearrangement of the allyl group to the neighboring carbon on the benzene ring, giving an allyl-alkoxyphenol (labelled as 4a, 4b or 4c). The latter compounds were alkylated to give the allyl-dialkoxy-substituted benzenes labelled as the 5-series (Paduraru et al., 2008). A few compounds were derived from eugenol (4-allyl-3-methoxyphenol) by alkylation, as described in Paduraru et al. (2008). para-Substituted compounds with a nitrogen atom on the benzene ring (3c{O2,N3}, 3c{N2,O3}) were derived from 4-aminophenol by bis-acetylation, with a carboxylic acid chloride or anhydride of the chain length desired on N, followed by saponification of the ester. The phenol group was then alkylated as described previously and, finally, the carboxamide was reduced with lithium aluminum hydride to the alkyl amine (Yang Yu, unpublished result). 175 Appendix B. Test compounds and their R groups used in a no-choice feeding bioassay on T. castaneum. 3a = ortho, 3b = meta, 3c = para substitutions, alkyl (me, et, pr, n-bu, i-pent). Individual compounds have a single R1 and a single R2 group; small libraries are blends of 5 compounds with a single R1 and a range of R2 groups. Treatment R1 R2 Compound name Control -- -- -- 3a{1,1-5} alkyl me 1-alkoxy-2-methoxybenzene 3a{2,1-5} alkyl et 1-alkoxy-2-ethoxybenzene 3a{3,1-5} alkyl pr 1-alkoxy-2-propoxybenzene 3a{4,1-5} alkyl n-bu 1-alkoxy-2-butoxybenzene 3a{6,1-5} alkyl allyl 1-alkoxy-2-allyloxybenzene 3b{1,1-5} alkyl me 1-alkoxy-3-methoxybenzene 3b{2,1-5} alkyl et 1-alkoxy-3-ethoxybenzene 3b{3,1-5} alkyl pr 1-alkoxy-3-propoxybenzene 3b{4,1-5} alkyl n-bu 1-alkoxy-3-butoxybenzene 3b{5,1-5} alkyl i-pent 1-alkoxy-3-isopentyloxybenzene 176 3b{2,2} et et 1,3-diethoxybenzene 3b{3,3} pr pr 1,3-dipropoxybenzene 3b{4,4} n-bu n-bu 1,3-dibutoxybenzene 3b{5,5} i-pent i-pent 1,3-di-isopentyloxybenzene 3b{6,6} allyl allyl 1,3-diallyoxybenzene 3b{3,5} pr i-pent 1-isopentyloxy-3-propoxybenzene 3b{1,5} me i-pent 1-methyl-3-isopentyloxybenzene 3b{1,6} me allyl 1-allyloxy-3-methoxybenzene 3b{2,6} et allyl 1-ethoxy-3-methoxybenzene 3b{3,6} pr allyl 1-allyloxy-3-propoxybenzene 3b{4,6} bu allyl 1-allyloxy-3-butoxybenzene 3b{5,6} i-pent allyl 1-allyloxy-3-isopentyloxybenzene 3c{1,1-5} alkyl me 1-alkoxy-4-methoxybenzene 3c{2,1-5} alkyl et 1-alkoxy-4-ethoxybenzene 3c{3,1-5} alkyl pr 1-alkoxy-4-propoxybenzene 3c{4,1-5} alkyl n-bu 1-alkoxy-4-butoxybenzene 3c{5,1-5} alkyl i-pent 1-alkoxy-4-isopentyloxybenzene 3c{1,1} me me 1,4-dimethoxybenzene 3c{2,2} et et 1,4-diethoxybenzene 3c{3,3} pr pr 1,4-dipropoxybenzene 3c{4,4} n-bu n-bu 1,4-dibutoxybenzene 3c{5,5} i-pent i-pent 1,4-di-isopentyloxybenzene 3c{6,6} allyl allyl 1,4-diallyloxybenzene 3c{1,6} me allyl 1-allyloxy-4-methoxybenzene 3c{3,4} pr n-bu 1-butoxy-4-propoxybenzene 3c{3,5} pr i-pent 1-isopentyloxy-4-propoxybenzene 3c{3,6} pr allyl 1-allyloxy-4-propoxybenzene 177 3c{5,6} i-pent allyl 1-allyloxy-4-isopentyloxybenzene 3a{N3,O6} pr allyl N-propyl-2-allyloxyaniline 3b{N2,O6} et allyl N-ethyl-3-allyloxyaniline 3b{N4,O3} n-bu pr N-butyl-3-propoxyaniline 3b{O5,N3} pr n-pent N-propyl-3-pentoxyaniline 3c{N2,O1} et me N-ethyl-4-methoxyaniline 3c{N2,O2} et et N-ethyl-4-ethoxyaniline 3c{N2,O3} et pr N-ethyl-4-propoxyaniline 3c{N2,O6} et allyl N-ethyl-4-allyloxyaniline 3c{N3,O3} pr pr N-propyl-4-propoxyaniline 3c{N3,O6} pr allyl N-propyl-4-allyloxyaniline 3c{O1,N3} pr me N-propyl-4-methoxyaniline 3c{O2,N3} pr et N-propyl-4-ethoxyaniline 3c{O4,N3} pr n-bu N-propyl-4-butoxyaniline 3c{O5,N3} pr n-pent N-propyl-4-pentoxyaniline 178 Appendix C. Percent feeding by insects on flour disk treated with four doses of various compounds after one and two weeks. The control run with each treatment was considered to be 100%. Mortality for each treatment was calculated by dividing the number of adult T. castaneum alive at the end of two weeks divided by the number original on each plate. Treatment 1 3a{1,1-5} % Feeding after one week % Feeding after two weeks Survival (# alive 14 days/beetles day 0) Dosage (µg/cm2) Dosage (µg/cm2) Dosage (µg/cm2) 50 100 200 1 50 100 200 1 50 100 200 105.1 101.8 92.3 90.8 106.2 98.6 96.6 93.4 1.0 0.96 1.0 1.0 3a{2,1-5} 95.9 90.3 108.6 -- 100.5 90.1 114.8 -- 1.0 1.0 1.0 -- 3a{3,1-5} 82.2 94.1 98.5 -- 85.3 95.8 97.8 -- 0.92 1.0 1.0 -- 3a{4,1-5} 98.5 96.7 96.7 -- 98.1 101.8 101.1 -- 1.0 1.0 1.0 -- 3a{6,1-5} 94.4 104.1 109.7 -- 100.5 105.4 113.6 -- 1.0 1.0 1.0 -- 3b{1,1-5} 91.5 98.0 90.8 -- 96.1 98.7 95.5 -- 1.0 0.96 0.92 -- 3b{2,1-5} 93.3 97.1 97.3 -- 97.2 99.4 95.9 -- 1.0 1.0 1.0 -- 3b{3,1-5} 102.6 105.2 -- -- 107.9 105.2 -- -- 1.0 1.0 -- -- 3b{4,1-5} 100.6 103.9 92.1 -- 104.2 105.7 102.5 -- 1.0 1.0 1.0 -- 3b{5,1-5} 104.7 106.5 105.4 94.4 104.5 103.8 109.5 112.2 1.0 1.0 1.0 1.0 3b{2,2} 21.4 64.3 0.0 -- -- -- -- -- 1.0 1.0 0.88 -- 3b{3,3} 21.4 42.9 171.4 -- -- -- -- -- 1.0 1.0 1.0 3b{4,4} 107.9 91.7 103.2 102.7 101.7 95.6 106.1 101.6 1.0 1.0 1.0 1.0 3b{5,5} 86.3 96.9 96.6 81.5 90.7 102.1 100.8 97.1 1.0 1.0 1.0 1.0 3b{6,6} 108.3 104.9 80.4 -- 106.2 85.7 105.4 -- 1.0 1.0 1.0 -- 3b{1,5} 116.7 -72.2 161.1 -- 53.2 162.0 159.5 -- 1.0 1.0 1.0 -- 3b{1,6} 52.8 27.8 27.8 -5.6 81.0 15.2 53.2 210.1 1.0 1.0 1.0 1.0 3b{3,5} 150.0 128.6 21.4 42.9 -- -- -- -- 1.0 0.88 1.0 0.68 3b{2,6} 123.2 74.6 90.8 -- 98.6 92.1 132.9 -- 1.0 1.0 1.0 -- 3b{3,6} 97.3 84.3 97.3 113.5 107.1 98.6 126.4 154.3 1.0 1.0 1.0 1.0 3b{4,6} 94.1 87.6 113.5 123.2 122.1 173.6 182.1 287.1 1.0 1.0 1.0 1.0 3b{5,6} 99.6 97.4 91.5 73.6 106.2 102.1 103.7 91.1 1.0 1.0 1.0 1.0 179 3c{1,1-5} 106.8 101.5 115.4 106.2 103.1 101.6 113.0 102.4 1.0 1.0 1.0 1.0 3c{2,1-5} 106.3 95.0 86.8 103.4 103.3 100.3 94.4 110.7 1.0 1.0 1.0 1.0 3c{3,1-5} 96.7 96.4 -- -- 104.2 102.8 -- -- 1.0 1.0 -- -- 3c{4,1-5} 101.5 95.3 104.7 -- 103.2 98.4 108.5 -- 1.0 1.0 1.0 -- 3c{5,1-5} 104.9 124.7 104.3 -- 105.5 118.0 -- -- 1.0 1.0 1.0 -- 3c{1,1} 115.1 120.2 113.4 96.9 101.5 100.7 108.9 93.6 1.0 1.0 1.0 1.0 3c{2,2} 115.5 -- -- -- 111.0 -- -- -- 1.0 -- -- -- 3c{3,3} 107.8 102.6 -- -- 114.1 105.4 -- -- 1.0 1.0 -- -- 3c{4,4} 99.5 97.1 29.8 42.1 107.2 99.0 83.6 48.9 1.0 1.0 1.0 1.0 3c{5,5} -- 93.1 102.0 102.6 -- 98.3 98.1 97.6 -- 1.0 1.0 1.0 3c{6,6} 101.7 93.6 92.2 -- 118.0 102.7 102.7 -- 1.0 1.0 1.0 -- 3c{1,6} 92.0 103.6 96.4 -- 96.1 102.9 98.5 -- 1.0 1.0 1.0 -- 3c{3,4} 97.4 88.5 104.7 79.1 98.9 93.8 109.2 100.1 1.0 1.0 1.0 1.0 3c{3,5} 102.6 82.9 100.7 89.1 101.5 85.6 106.6 104.1 1.0 0.96 1.0 1.0 3c{3,6} 105.3 109.5 97.9 97.6 101.8 105.9 100.5 104.8 1.0 1.0 1.0 1.0 3c{5,6} 110.4 101.3 105.7 -- 106.4 100.3 102.5 -- 1.0 1.0 1.0 -- 3a{N3,O6} 87.5 92.3 97.1 103.9 88.3 97.9 98.4 106.5 0.96 1.0 1.0 1.0 3b{N2,O6} 94.1 103.2 100.7 -- 93.8 105.2 99.3 -- 1.0 1.0 1.0 -- 3b{N4,O3} 106.2 97.0 97.3 93.8 106.6 103.3 110.8 96.7 1.0 1.0 1.0 1.0 3b{O5,N3} 92.8 90.3 91.4 110.4 94.2 84.5 96.1 119.7 1.0 1.0 1.0 1.0 3c{N2,O1} 116.5 97.9 93.1 123.2 112.1 88.2 97.0 109.2 1.0 1.0 1.0 1.0 3c{N2,O2} 101.5 99.8 109.6 113.8 95.6 100.2 109.1 105.9 1.0 1.0 1.0 1.0 3c{N2,O3} 97.3 101.1 92.2 -- 102.7 109.5 100.2 -- 1.0 0.96 1.0 -- 3c{N2,O6} 105.5 112.4 108.9 113.8 111.4 110.0 111.6 109.9 1.0 1.0 1.0 1.0 3c{N3,O3} 115.1 109.4 98.6 101.1 110.8 113.6 109.1 102.9 1.0 1.0 1.0 1.0 3c{N3,O6} 100.3 109.4 107.9 100.8 95.8 105.5 107.5 102.5 1.0 1.0 1.0 0.84 3c{O1,N3} 102.7 105.0 110.5 120.7 101.9 102.1 106.6 115.9 1.0 1.0 1.0 0.96 3c{O2,N3} 93.4 104.0 99.8 104.0 94.5 103.9 100.7 105.8 1.0 1.0 1.0 1.0 3c{O4,N3} 96.6 99.1 112.5 112.8 102.2 103.4 107.8 104.2 1.0 1.0 1.0 1.0 3c{O5,N3} 99.2 108.2 102.0 89.8 100.7 104.6 102.7 95.7 1.0 1.0 1.0 1.0 180 181