Competitive interactions among two specialist predators, <i>Laricobius nigrinus</i> Fender (Coleoptera: Derodontidae) and <i>Sasajiscymnus tsugae</i> Sasaji and McClure (Coleoptera: Coccinellidae), and a generalist predator, <i>Harmonia axyridis</i> Pallas (Coleoptera: Coccinellidae), of hemlock woolly adelgid were evaluated using laboratory, field and video studies. The two specialist predators are part of a biological control program for <i>A. tsugae</i>, and the potential for competition among these species and previously established generalist predators is unknown. In laboratory studies of predator groups in Petri dish assays, the only significant negative effects from competition occurred among conspecifics, resulting in reduced net egg production by <i>L. nigrinus</i> and <i>H. axyridis</i> and reduced feeding by <i>H. axyridis</i>. In contrast, heterospecific combinations showed non-interference. In longer duration field studies of predator groups, held in branch enclosures, predator survival and feeding were not significantly affected by additional predators. Net reproduction was again significantly reduced by conspecifics, while heterospecifics showed non-interference for all predator responses. All predators reduced the number of <i>A. tsugae</i> nymphs of the next generation relative to no-predator controls; however, <i>L. nigrinus</i> had much greater impact overall due to the large number of progeny produced. Video studies revealed that predator behavior varied qualitatively and quantitatively by species, and did not appear to be coordinated temporally or spatially. All species exhibited continuous activity patterns that were punctuated by longer periods of rest. The specialist predators were more selective of feeding and oviposition sites, and rested at more concealed locations than <i>H. axyridis</i>. Conspecifics significantly altered the time allocated to specific behaviors for <i>L. nigrinus</i> and <i>H. axyridis</i>, resulting in reduced predator effectiveness due to increased searching and decreased feeding and oviposition. All predator groups maintained a high degree of spatial separation relative to assay size, suggesting that chemical or tactile cues may be used to regulate their distributions. Overall, these studies suggest that the three predator species will be compatible in this system. Management implications include using multiple-predator species combinations over single-species for biological control of <i>A. tsugae</i> and implementing low-density releases to reduce the potential negative effects of intraspecific competition. / Ph. D.

When multiple predators foraging together have different individual consumption rates than predators foraging in isolation, they exhibit non-independent multiple predator effects on prey. I examined multiple predator effects in a system consisting of invasive green crabs (Carcinus maenas L.), native rock crabs (Cancer irroratus Say) and benthic macrofauna prey. First, I examined multiple predator effects when green crabs and rock crabs forage on soft-shell clams (Mya arenaria L.) in different habitat types (sand, sand with artificial seagrass) and assessed the behavioural mechanisms responsible for the observed predation effects. Independent multiple predator effects on prey were detected for most conspecific and heterospecific pairs in both habitat types. In general, crab foraging behaviours were not affected by the presence of another predator. Interactions between predators did not influence foraging behaviours because encounters were infrequent, short in duration and predominantly non-aggressive. A non-independent multiple predator effect on prey (marginally significant) was observed when green crabs foraged with rock crabs in artificial seagrass. This effect, however, could not be explained by the observed crab behaviours. Second, I investigated multiple predator effects when green crabs and rock crabs forage on a soft-sediment macrofauna community. Because crabs did not have significant predation effects on the community throughout the experiment, I did not evaluate multiple predator effects on prey. It is possible that crab predation was not important in regulating the macrofauna community, in which case multiple predator effects were non-existent. Predation may have been suppressed due to a combination of factors, including interactions between predators, harsh environmental conditions or a sub-optimal prey field. Alternatively, my ability to detect significant predation effects may have been hindered because of prey movement in and out of cages or low statistical power. Overall, results from this thesis demonstrate that multiple predator effects on prey may differ with habitat and highlights the importance of conducting behavioural observations to better understand interactions between predators and the resulting consequences for prey. Multiple predator effects on a soft-sediment community should be re-evaluated to assess the importance of these crab species in regulating benthic macrofauna under natural conditions.

Dans cette thèse nous regardons plusieurs modèles de populations structurés s’écrivant à l’aide d’équations de transport. Le caractère bien posé ainsi que la positivité des solutions sont montrés de manière systématique au sens des sémiologues dans un cadre L1. Un premier travail est consacré à un système de type proie prédateur structuré en âge. Une étude de stabilité des équilibres nous permet de formuler explicitement un seuil un seuil d’extinction ainsi qu’in seuil pouvant amener à l’explosion des populations. On obtient numériquement la possibilité d’un cycle limite ainsi que la convergence vers un équilibre de coexistence des populations. Dans un cas particulier, ce modèle se réécrit comme un système différentiel à retard. A l’aide de fonctionnelle de Lyapunov, on montre la stabilité globale de cet équilibre sous certaines conditions. On étudie également 2 modèles structuré en taille, issus de la dynamique cellulaire. L’un est composé de deux équations de transport où la cellule peut être soit prolifèrent soit quiescente ; et le deuxième est une équation de type transport/ diffusion avec des conditions aux bords FELLER. On vérifie à chaque fois l’irréductibilité du semi groupe puis des arguments de faibles capacité L1 nous donne l’existence d’un « gap spectral » sous certaines conditions. On démontre ainsi dans certains cas la croissance exponentielle asynchrone du semi groupe / This thesis is dedicated to some structured populations models described with transport or transport-diffusion equations. The well-posedness, in the semigroupes setting in L1 and the positivity of the solutions are systematically shown. A first work is dedicated to an age-structured predator/prey system. A stability study of the equilibria allow us to give explicit formulations of an extinction threshold and an threshold which can lead to explosion of solutions. We numerically obtain the possibility to get a limit cycle and the convergence to a coexistence equilibrium of the populations. In a specific case, this model rewrites as a delay differential system. Using Lyapunov functional, we show the global stability of this equilibrium under some assumptions. We also study two size-structured models that come from cellular dynamics. The first one consists on two transport equations, where the cell can either proliferate or be quiescent, and the second one is a transport-diffusion equation with Feller boundary conditions. The irreducibility of the semigroup governing this latter model is always satisfied using the Hopf maximum principle. However, the irreducibility for the first model is true only under a necessary and sufficient condition that we give. We also show for these two models, using some weak compactness arguments in L1, the existence of a `spectral gap' (essential type strictly less than the type) ensuring the asynchronous exponential growth of the semigroup.