UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Muddy molecules for pandemic protection : investigating the use of wetland sediment as a tool for the surveillance of avian influenza virus in wild waterfowl birds Coombe, Michelle Karin


Avian influenza virus (AIV) is a wildlife-origin pathogen with potential for causing disease in multiple species but challenges in sampling directly from wild bird hosts means the diversity of circulating viruses is often incompletely understood. This dissertation evaluated if indirect sampling of wild waterfowl birds through viruses excreted into their wetland environment—combined with probe capture sequencing—could provide an alternative to conventional AIV surveillance. Our project collected both conventional bird swabs (n = 1,290) and environmental wetland sediment samples (n = 2,090) between September 2016 and August 2018 in southwestern British Columbia, Canada. To facilitate selection of environmental sample collection locations, we characterized the spatial clustering of AIV positivity in wetland using Kulldorff's spatial scan statistic and identified ecological variables associated with detection of AIV through a Bernoulli generalized additive regression model with the R-INLA program. Evaluating the use of environmental samples as a proxy for conventional surveillance was done by comparing the AIV subtype diversity found in wetland sediment with that from direct bird sampling. The temporal dynamics of AIV subtype diversity in wetland sediment follows the seasonality of waterfowl migration. Furthermore, it suggests that extended persistence of AIV in the wetland sediment was possible but unlikely to extend over the summer months in our study area. Compared to direct bird surveillance, environmental surveillance provided higher subtype detection rates, detected subtypes at more times of the year, and detected a higher maximum number of subtypes per sample. We developed a methodology to incorporate both the fragmented RNA detected in the environment and complete AIV sequences from wild birds into the same Bayesian phylogenetic trees. Within a subtype, we found that only a small number of AIV sequences from sediment were considered genetically equivalent to AIV from study birds; however, many were genetically similar. Overall, this dissertation demonstrates that while environmental surveillance can provide similar trends to conventional surveillance, greater genetic diversity is detected at more locations and more times of year when using both programs together. Combining multiple surveillance data sources provides the best context for decision makers in a complex and incompletely understood disease landscape.

Item Media

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International