Intraguild (IG) predation, where species within the same guild prey on each other, is common in aquatic communities. I used the abundance and distributional patterns of three species of Rhyacophila (Trichoptera), derived from a survey of 25 streams in Nova Scotia, Canada, to test several predictions of IG theory. I first sampled microhabitats and conducted behavioural observations to establish that the species do co-occur and prey on each other. Abundance patterns did not conform to two key model predictions: (1) Neither of the IG prey (R. minor, R. vibox) was excluded from the most productive streams, and (2) IG prey densities were not inversely related to productivity in streams with the IG predator (R. fuscula). It remains possible that intraguild predation occurs between these species of Rhyacophilia, the measured abundances do not indicate any measurable effects from IGP, the abundance being determined instead by the availability of prey.

The thesis examines the dynamics of intraguild predation models incorporating adaptive behaviour. The top predator varies its relative consumption of the intermediate consumer and the basal resource. The consumer alters its activity level in response to the threat of predation. Incorporating these adaptive behaviours facilitated three species coexistence, but restricted omnivory by promoting the formation of three species food chains. Model modifications which promoted omnivory also tended to reduce three species coexistence. It is predicted that omnivory should be most common when the intermediate and the basal species are similarly profitable to the predator. The model also predicts two types of omnivory. Strong omnivory occurs when the predator always consumes intermediate amounts of both prey items. Weak omnivory occurs when the predator preys almost exclusively on one prey at most times, but rapidly switches to include the other when the relative abundance of that second prey exceeds a certain level.

The thesis examines the dynamics of intraguild predation models incorporating adaptive behaviour. The top predator varies its relative consumption of the intermediate consumer and the basal resource. The consumer alters its activity level in response to the threat of predation. Incorporating these adaptive behaviours facilitated three species coexistence, but restricted omnivory by promoting the formation of three species food chains. Model modifications which promoted omnivory also tended to reduce three species coexistence. It is predicted that omnivory should be most common when the intermediate and the basal species are similarly profitable to the predator. The model also predicts two types of omnivory. Strong omnivory occurs when the predator always consumes intermediate amounts of both prey items. Weak omnivory occurs when the predator preys almost exclusively on one prey at most times, but rapidly switches to include the other when the relative abundance of that second prey exceeds a certain level.

Commercial fishing in the Baltic Sea has led to decrease in abundance of large predatory fish and as a result the predation pressure on smaller fish species has been reduced. Three-spined stickleback (Gasterosteus aculeatus) is among the species that have benefited from reduced predation pressure. Sticklebacks are a small fish species that spends its juvenile stage in coastal habitat and its adult stage in the open sea with a yearly migration from the sea to coastal habitat and lakes for spawning. Sticklebacks have been shown to have a negative effect on fry of coastal species, such as perch (Perca fluviatilis). Therefore an increase in stickleback abundance, as a consequence of reduced predation pressure, can potentially further reduce abundance of fry of coastal species, such as perch. On the other hand sticklebacks could be subject to predation from adult perch. This two-species interaction is studied in the perspective of general ecological IGP (intra guild predation) theory. In order to determine how the observed increase of three-spined sticklebacks negatively or positively affects the growth of different stages of perch, a time series of operculum bones from perch collected by the coastal monitoring program prior to the increase of sticklebacks and up until present was analyzed. Three areas along the Swedish coast showing increased stickleback abundance (Gaviksfjärden, Norrbyn and Holmön) were compared to areas that had no sticklebacks or only a small increase in abundance (Kinnbäcksfjärden, Råneåfjärden, Långvindsfjärden) prior and after the general increase in stickleback density. In addition to the growth analysis, a dietary analysis of perch was carried out in order to determine to what extent perch utilize sticklebacks as a food source. I found that there were significant differences in regard to perch growth between the two time periods; however there was no significant difference in growth between the control areas and the stickleback areas except for perch of medium size. Further, this difference in growth between the control and stickleback areas was present both before and after the stickleback increase and could therefore not be tied directly to stickleback increase. Consequently, results from a multiple regression showed that stickleback abundance was not able to significantly explain the observed changes in perch growth. The dietary analysis showed that larger and intermediate perch tend to consume sticklebacks as well as benthic fauna and other fish species while smaller perch tended to feed exclusively on smaller invertebrates. This study concluded that the increase in stickleback abundance had only a very weak effect on perch growth.

A model of a three species intraguild predation community is proposed. The model is realized as a system of cross-diffusion equations which allow the intraguild prey species to adjust its motility based on local resource and intraguild predator densities. Solutions to the cross-diffusion system are shown to exist globally in time and the existence of a global attractor is proved. Abstract permanence theory is used to study conditions for coexistence in the ecological community. The case where the intraguild prey disperses randomly is compared to the case where the intraguild prey disperses conditionally on local ecological fitness and it is shown that the ability of the intraguild prey to persist in the ecological community is enhanced if the intraguild prey utilizes a movement strategy of avoiding areas with negative fitness. A finite element scheme is used to numerically simulate solutions to the system and confirm the analytical results.

Coexistence among ecologically similar species is often facilitated through temporal or spatial partitioning mechanisms that reduce or eliminate direct interaction. However, in many communities exhibiting guild structure, wherein potential competitors may also prey on one another, sympatric relationships persist despite species' similar life history strategies, spatial and temporal restrictions imposed by ephemeral habitats, and resource limitations that promote competition and predation. To identify the ecological roles of species-specific behavioral patterns within aquatic guilds, I quantified larval intraspecific agonistic behavior among two species of intraguild (IG) predators, Ambystoma opacum and A. tigrinum, and their shared intraguild prey, A. maculatum. All species exhibited similar ontogenetic patterns of aggression, characterized by peaks of aggression early in development and subsequent gradual decreases through metamorphosis. However, the intensity of aggression varied considerably among guild species through development, as did behavioral responses to varying levels of ambient water temperature, invertebrate prey density, and presence of predatory odonate naiads. The observed patterns suggest that guild species, despite morphological and physiological similarities, exhibit unique behavioral responses through ontogeny and in response to habitat variables, suggesting that temporally staggered breeding phenologies have contributed to behavioral divergence among these sympatric congeners. However, in situ observations of larval behavior, although largely in agreement with laboratory results on timing of increased aggression, indicated that IG predators exhibited pond-level species partitioning and do not necessarily co-occur despite being regarded as sympatric. These results, taken together with observed species-specific impacts of IG predators on IG prey, suggest that ecologically similar IG predators exert widely differing predatory pressure on shared prey, and that similarities among guild species may ultimately result in habitat partitioning across local scales.

One of the basic questions of concern in mathematical biology is the long-term survival of each species in a set of populations. This question is particularly puzzling for a natural system with omnivory due to the fact that simple mathematical models of omnivory are prone to species extinction. Omnivory is defined as the consumption of resources from more than one trophic level. In this work, we investigate three omnivory models of increasing complexity. We use the notion of permanent coexistence, or permanence, to study the long-term survival of three interacting species governed by a mixture of competition and predation. We show the permanence of our models under certain parameter restrictions and include the biological interpretations of these parameter restrictions. Sensitivity analysis is used to obtain important information about meaningful parameter data collection. Examples are also given that demonstrate the ubiquity of omnivory in natural systems. / Ph. D.

This thesis outlines research conducted as part of a collaborative project between Virginia Tech and Plant and Food Research New Zealand (PFRNZ) to introduce Heringia calcarata (Loew) (Diptera: Syrphidae) to New Zealand (NZ) for biological control of woolly apple aphid (WAA), Eriosoma lanigerum (Hausmann) (Hemiptera: Aphididae). Ultimately, the introduction of H. calcarata to New Zealand will be contingent upon satisfying regulatory requirements and concerns, including documentation that it will not have an adverse effect on the existing biological control of WAA by Aphelinus mali (Haldeman) (Hymenoptera: Aphelinidae). As well, it will be critical to develop methods for sustained rearing of H. calcarata in captivity. Basic and applied studies were conducted toward providing essential information for advancing this project. Apple shoot sections with a WAA colony that did or did not contain mummified aphids parasitized by A. mali were deployed in pairs at the base of apple trees. There was no significant difference in the mean number of H. calcarata eggs deposited between shoots with parasitized (1.5 ± 0.34 SE) and non-parasitized colonies (1.75 ± 0.42 SE), although female H. calcarata laid eggs less frequently on colonies with a high percentage parasitization. In choice-test feeding studies, larvae were offered non-parasitized aphids in combination with aphids in an early stage of parasitization or mummified aphids. Larvae consumed significantly fewer aphids in an early stage of parasitization (10.8 ± 0.48 SE) than non-parasitized aphids (13.4 ± 0.42 SE) and very few mummies (0.4 ± 0.14 SE) compared with non-parasitized aphids (14.2 ± 0.4 SE). In no-choice feeding trials, larvae consumed significantly more non-parasitized aphids (25.3 ± 1.93 SE) than aphids in an early stage of parasitization (19.7 ± 1.85 SE) or mummified aphids (2.2 ± 0.71 SE) and significantly fewer mummified aphids were consumed than early parasitized aphids. WAA colonies in situ on the branches of potted apple trees were exposed to H. calcarata, A. mali, or both. Exposure to H. calcarata larvae independently and in combination with A. mali was shown to have a significant effect on the number of WAA compared with control colonies, and H. calcarata larvae did not affect the number of mummified aphids produced within colonies. Heringia calcarata eggs were collected by deploying excised apple shoot sections containing at least one WAA colony at the base of apple trees for 8-12 h. One or more eggs were laid on 29% of shoots (n = 233 shoots). On shoots with eggs, 2.4 ± 0.21 SE eggs per shoot were recorded. In 2012, four shipments of H. calcarata eggs and larvae (total of 178) were sent from Virginia to a quarantine containment facility in NZ. This demonstrated that juvenile H. calcarata could be successfully transported internationally. In total, 124 adult flies were generated in NZ, representing 69.9% of the number of eggs and larvae recovered upon delivery to quarantine. Field-collected gravid female H. calcarata oviposited on WAA colonies under captive conditions: 63% in 2011 (n = 8) and 80% in 2012 (n = 15). In 2012, 98% of the eggs deposited hatched. Virgin females reared from eggs in the laboratory developed mature oocytes regardless of access to pollen. The findings of this research offer valuable insights into the biology and ecology of H. calcarata that are directly relevant to the project goals and that will help guide the development of H. calcarata as a classical biological control agent for WAA in NZ. / Master of Science in Life Sciences

Traditional views of food web dynamics have characterized species interactions as linear and direct. However, modern food web theory suggests that interactions can also be nonlinear and indirect, so that disturbance at one trophic level is transmitted throughout the community. Many previous studies have demonstrated that the removal of top predators from terrestrial ecosystems can have broad-scale impacts on community ecology. I examined the direct and indirect effects of raccoon removal from sea turtle nesting beaches in east-central Florida during the summers of 2003 and 2004. Raccoon and ghost crab predation are among the highest causes of egg mortality for sea turtles in Florida and raccoons are intraguild predators of ghost crabs. Because of the damage done to sea turtle nests, raccoons have been removed from some beaches during the sea turtle nesting season. I compared the diet and demography of a raccoon population that had experienced two decades of raccoon removal to a population that had previously been unmanipulated. I found that long-term raccoon removal had created a significantly male-biased sex ratio in that population. I also examined the indirect effects of raccoon removal by comparing the abundances of raccoons and ghost crabs at four study sites using passive tracking plots. My data suggest intraguild predation by raccoons limits ghost crab abundance and that reduced raccoon abundance allowed ghost crab abundance to increase, resulting in a net increase in sea turtle egg predation. These results support my hypothesis that intraguild predation of ghost crabs by raccoons is an influential interaction on sea turtle nesting beach community dynamics.

Biotic interactions are known to shape natural community assemblages and biodiversity. Positive interactions such as facilitation have recently received attention in ecological food webs. Mechanistic models have improved our understanding of these complex food web interactions. Here, focus is given to a three-species food web system with a beach dune natural community in mind. In the last decade, there has been a series of studies investigating intraguild predation between two major loggerhead sea turtle nest predators, North American raccoons and Atlantic ghost crabs. Studies have also highlighted that ghost crab predation assists raccoons in finding nests (i.e., facilitated predation). However, the combined effects of these two intraguild interactions and their consequences on nests have not been examined explicitly. The aims of this study were to (i) develop a three-species, ordinary differential equation model (ii) implement a sensitivity analysis to understand the influence of facilitation and other factors in driving species richness and abundance and (iii) characterize the dynamic interactions between intraguild predators and their effects on a shared resource. Interactions between ghost crabs and sea turtle eggs and facilitation can yield a wide variety of species abundance responses and were influential factors in the model. I found that high secondary sea turtle egg depredation and low facilitated predation by raccoons led to three species co-existence regions in the model. Controlling for nest predators at higher abundance levels showed that ghost crabs had a larger negative effect on sea turtle egg abundance responses when compared to raccoons. This suggests that interactions between sea turtle eggs and ghost crabs appear to be important and potential sea turtle nest management implications are discussed such as the use of ghost crab exclusion devices.