The occurrence and ecology of Toxoplasma gondii in a terrestrial arctic food web

Abstract

The occurrence and ecology of the apicomplexan parasite Toxoplasma gondii in arctic wildlife is not well understood. Transmission cycles, especially in terrestrial systems, are enigmatic because environmentally resistant oocysts, shed by felid definitive hosts, might be less responsible for transmission in the Arctic than in more southern latitudes. Toxoplasma gondii can also be transmitted through the food web by carnivory, and by transmission from mother to fetus during gestation, and these routes are thought to play a large role in the ecology of T. gondii in the Arctic. In this thesis, I examine T. gondii in a well-described part of the food web at Karrak Lake, Nunavut, in the central Canadian Arctic and through experimental infections of domestic waterfowl. In the field over 3 years, I sampled generalist carnovires (arctic foxes), migratory herbivores (Ross’s Geese and Lesser Snow Geese), and resident herbivores (lemmings). Using an occupancy modeling approach that accounted for imperfect detection, I compared commonly used serological assays to estimate prevalence of T. gondii antibodies in sera from arctic foxes and eluted blood on filter paper from Ross’s geese and Lesser Snow Geese and compared commonly used serological assays. I also used a naïve estimator to determine prevalence of T. gondii antibodies in sera from Ross’s Geese and Lesser Snow Geese, and blood on filter paper from lemmings. I detected antibodies against T. gondii in sera from arctic foxes (47-60%, depending on age category), Ross’s Geese (32%) and Lesser Snow Geese (28%). I also detected antibodies in blood on filter paper from Ross’s Geese (39% seropositive) and Lesser Snow Geese (36% seropositive) but not in lemmings. These findings suggest that light geese might introduce T. gondii to the Karrak Lake ecosystem with the annual spring migration and that oocyst transmission might not occur in the terrestrial system, because the parasite was not detected in resident rodents. For the in vivo experimental infections, we used a novel application of a multi-scale occupancy framework to determine within-host detection probability of T. gondii in experimentally inoculated domestic geese and then used those results to guide tissue sampling in wild Ross’s Geese and Lesser Snow Geese. In the experimental inoculation trial, the heart and brain had the highest detection probability for T. gondii through a real-time PCR with melt-curve analysis. Toxoplasma gondii DNA was not detected in tissues from wild geese, suggesting that the parasite was either not present, or methodological difficulties prevented its detection. The research presented in this thesis forms the groundwork for further T. gondii studies in this region.