Aggregation, dispersal and edge behavior of the bark beetle predator Thanasimus dubius (Coleoptera: Cleridae); potential use in biological control

Author

Arnaud Costa, Southern Illinois University CarbondaleFollow

Date of Award

5-1-2010

Degree Name

Doctor of Philosophy

Department

Zoology

First Advisor

Reeve, John

Abstract

Bark beetles are major pests of pine forests in North America that can inflict considerable damage and cause severe economic loss. The checkered beetle Thanasimus dubius Fabricius (Coleoptera: Cleridae), an abundant predator in the Eastern USA and Canada, has been suggested to influence the dynamics of several bark beetles, including the economically important southern pine beetle (SPB), Dendroctonus frontalis Zimmermann (Coleoptera: Curculionidae). Because of the importance of T. dubius in pine forests, I studied several aspects of T. dubius biology including both its chemical and dispersal ecology. I first present an experiment that was conducted to improve the rearing system of T. dubius so that sufficient numbers of predators could be produced more efficiently for both behavioral and ecological studies. Improvements to the rearing method potentially could be achieved by adding a preservative to increase the shelf-life of the diet used to feed larval T. dubius thereby allowing feeding intervals to be extended. To evaluate this, I added sorbic acid, a preservative, to the larval diet at three different concentrations (0, 0.1 and 0.2%) and for two different intervals between feedings (2-3 vs. 5 days). Additionally, I measured potential deleterious effects of this preservative on predator performance (i.e., female fecundity). I then assessed the effect of this predator on SPB survival by releasing newly hatched T. dubius larvae at several densities (0, 50 100, and 200) on pine logs infested by SPB. I also studied the chemotactic response of this predator toward various volatiles such as prey pheromones or tree volatiles in a wind tunnel. I developed a method that associated upwind flight behavior and antennal behavior to quantify the relative attractiveness of each tested source of chemical. I then conducted three experiments to investigate the ability of individual T. dubius to respond to different prey olfactory signals (bark beetle pheromones) and assess their potential for learning. First, I examined individual predator preferences toward three major prey pheromones (ipsenol, ipsdienol, and frontalin) in a study that combined a mark-release-recapture experiment with a field choice test. I also measured the responses of individual wild T. dubius in a wind tunnel, which allowed the testing of a wider range of semiochemicals (frontalin, ipsenol, ipsdienol, sulcatol, and α-pinene). A final wind tunnel experiment tested the ability of naive predators to learn two prey pheromones, frontalin and ipsenol, that were presented alone or associated with a reward. In the last chapter, I quantified the dispersal of this predator and its behavior along an edge separating a pine forest from a non-suitable habitat. I analyzed T. dubius distribution and movement in the field relative to its bark beetle prey Ips grandicollis Eichhoff (Coleoptera: Curculionidae) and to the root feeder Dendroctonus valens LeConte (Coleoptera: Curculionidae). These three insects are associated with the phenomenon of red pine decline in the Great Lakes area. Thus, understanding the relative pattern of movements of these three beetles could provide considerable knowledge on the spatial and temporal progression of red pine decline by, for instance, assessing the connection between existing pockets. I estimated the dispersal quantiles for this predator relative to its bark beetle prey and determined whether dispersal behavior was relatively homogeneous (one kind of disperser) or heterogeneous (two kinds). In a second part, I sampled the abundance of T. dubius, I. grandicollis, and D. valens along transects set between a matrix area constituted by a clearing and a habitat zone composed of red pine forest. I further developed a diffusion model, including a constant k (corresponds to the ratio of densities along the edge), that permits characterization of beetle behavior around the boundary. Studies developed in the present dissertation, therefore, investigated several aspects of T. dubius ecology: predation on SPB, ability to respond to an olfactory signal, and dispersal traits. I first showed that augmentation with larval T. dubius can successfully reduce the SPB ratio of increase in infested pine logs. I also developed a system of rearing T. dubius that was more efficient and could allow larger numbers of predators to be produced. Sorbic acid did not reduce adult lifetime or size but did affect female fecundity (20 - 40% reduction). Increasing feeding time interval to 5 days (instead of 2-3 days) improved the efficiency of the system, even when sorbic acid was not added in the diet. Wind tunnel experiments present a methodology that could enable future quantification of the relative attractiveness of various semiochemical treatments. I identified a particular antennal behavior, which I called "stand up antennal" that was typically produced when the insect was attracted to a semiochemical. The choice tests in the field and the wind tunnel experiment using wild adults both showed that T. dubius individuals are generalists that are flexible in their response to a broad range of semiochemical signals (frontalin, ipsenol, ipsdienol, sulcatol, and α-pinene). I demonstrate that naive predators are more attracted toward ipsenol after having experienced conditioning with this kairomone via a reward. Finally, experiments on dispersal and edge behavior revealed that T. dubius and one of its common prey, the bark beetle I. grandicollis, exhibit a different pattern of movement within red pine stands and surrounding habitats. First, adult T. dubius can disperse substantially further (50% dispersed beyond 1.54 km) than I. grandicollis adults (50% dispersed beyond 0.13 km). Second, T. dubius was highly affected by the presence of an edge, showing a step-like pattern with very predators caught in unsuitable areas. Conversely, the behavior of I. grandicollis was less limited by such a boundary. As with T. dubius, the root colonizer D. valens showed a strong edge behavior that was characterized by avoiding matrix zones.