AN INSECT PHEROMONE CHANGES PLANT BEHAVIOR: LEVERAGING A SHARED SENSORY CUE FOR CROP PROTECTION

146 pagesAs demand for agricultural products rapidly grows, so too does the demand for effective biocompatible pest treatments. One area for novel treatments that has illustrated promise is the use of semiochemicals to alter the behavior of pest insects. By using these message-carrying compounds, growers can safely attract target insects to traps, repel them from areas under management, or simply confuse them. Currently, the semiochemicals used are typically derived from the pest insects such as pheromones, however insects detect and respond to a diversity of cues. Recent research has shown that the odors of predators can also elicit a strong response in prey, leading to questions about their use to protect crop plants such as; can the synthetic cues of predators be effectively deployed as a crop treatment? Even broader, semiochemical treatments can also provide important information for other trophic levels in a system such as the plants we are often attempting to protect. This multiple trophic effect highlights the potential of expanding semiochemical treatments beyond just the target pest. Many compounds in the biosphere are shared across trophic levels and this phenomenon may be exploited to both meet management goals and to improve our understanding of the sensory ecology of systems. The overarching goal for this dissertation was to learn how certain compounds affect multiple organisms in an agricultural setting and to develop and test the use of a synthetic predator pheromone treatment for reducing pest damage. To meet this goal, much had to be learned about the organisms involved and in doing so, unexpected and fascinating discoveries were made along the way. The study system involved Solanum tuberosum, the most damaging insect pest of S. tuberosum and other Solanum crops Leptinotarsa decemlineata (Say) (Colorado potato beetle), and their widely distributed predator, the Podisus maculiventris (Say) (spined soldier bug). Building on previous research in this system and after numerous trials, it was established that a large pheromone gland, the dorsal abdominal gland, contained within the predator was eliciting the strongest response in the beetle prey. In the first chapter, I set out to learn more about the development of the dorsal abdominal gland (DAG) and the release behavior of the gland contents. First, P. maculiventris were dissected at three adult stages and the chemical composition and amount of each compound was quantified. This revealed that the glands are not formed within 24 hours post eclosure and were mostly formed after seven days. After learning about the physiology of the gland development, the pheromone release behavior was investigated next. Through trials using VOC dataloggers, we found that no pheromone releases were detected in the newly eclosed adults and that the majority of the mature adult volatile chemical releases occurred in a scotophase pattern. These findings add new information about the development of exocrine glands in Pentatomidae and improve our understanding of how odors from the P. maculiventris modulate species interactions in cropping systems. The ephemeral properties of the DAG semiochemical also help explain some of the variability in treatment effect noted in past tests using live P. maculiventris as an odor source. In chapter 2, I field tested the use of live predators as an odor source and compared L. decemlineata response to a synthetic dorsal abdominal gland formulation in open dispersion release devices. During this preliminary work of deploying the treatments in the field, a period of time was spent piloting various release devices to find a functional, safe, and economical option to administer the synthetic P. maculiventris pheromone. Across two field seasons the results showed that feeding damage by L. decemlineata was 22 percent lower in predator odor treated plots, however the effect varied over the season, and in the second year the treatment effect was lost. Interestingly, the synthetic predator pheromone reduced plant damage more consistently than the live predator treatment. These results suggest that temporal patterns of predator cue release and strength may drive the prey’s response across the season, and that the synthetic pheromone dispensers may be a viable option to modify pest insect behavior in agricultural systems. Following the promising field results, in Chapter 3 I set out to better describe the mechanisms behind the prey response in order to optimize the synthetic predator dispensers. Most all organisms enact protective measures to reduce the chance of being consumed. Analogous to the concept of trophic cascade, we have generally considered the cues that are responsible for eliciting anti-predation behavior to follow a downward linear path, where the cues emanating from higher trophic levels cascade downward to prey. However, the role of basal trophic levels such as plants and their own ability to sense and respond to their environment has the potential to influence prey. I used lab and field experiments to ask if the predator pheromone alters plant quality and reduces the performance of L. decemlineata. I found evidence that the predator pheromone induces a defensive response in plants and reduces herbivore prey presence at all life stages and the amount the plant material they consume. The results expand our understanding of species interactions by considering the movement of sensory information, where the cues released from an insect predator are perceived by a plant, which has a negative effect on an herbivorous prey species. In light of these findings, we use the term indirect non-consumptive effects to describe predator-prey interactions where predatory sensory cues affect prey through another trophic level. Intrigued by the behavioral changes in the plant that were elicited by the predator pheromone, in Chapter 4 I fractionated the pheromone to better understand the plant response. Of the 5 primary compounds in the pheromone blend, 2 of the compounds are also shared with S. tuberosum as green leaf volatiles. Green leaf volatiles are released aerially from nearly all green plants when damaged that are then available for other portions of the plant and neighboring plants to detect and prepare for a potential damage agent via priming of plant defensive metabolites. My hypothesis was that the portion of the predator pheromone that is shared with plant green leaf volatiles was responsible for behavioral changes in the plant. The results supported the hypothesis, where the blend containing cues that are also shared with the plant were responsible for eliciting the greatest response in S. tuberosum and L. decemlineata. More broadly, this work emphasized how certain sensory cues in natural and managed systems are pervasive and used by multiple interacting trophic levels. I discuss the potential of applying shared cues, as opposed to single target organism-based treatments, to achieve a magnified effect beyond a single target trophic level