Ecological determinants of avian influenza virus, West Nile virus, and avian paramyxovirus infection and antibody status in blue-winged teal (Anas discors) in the Canadian prairies / Rodolfo Nallar, Zsuzsanna Papp, Frederick A. Leighton, Tasha Epp, John Pasick, Yohannes Berhane, Robbin Lindsay, and Catherine Soos.

Includes bibliographical references (pages 44-46).

The Canadian prairies are one of the most important breeding and staging areas for
migratory waterfowl in North America. Hundreds of thousands of waterfowl of numerous species
from multiple flyways converge in and disperse from this region annually; therefore this region
may be a key area for potential intra- and interspecific spread of infectious pathogens among
migratory waterfowl in the Americas. Using Blue-winged Teal (Anas discors, BWTE), which have
the most extensive migratory range among waterfowl species, we investigated ecologic risk factors
for infection and antibody status to avian influenza virus (AIV), West Nile virus (WNV), and avian
paramyxovirus-1 (APMV-1) in the three prairie provinces (Alberta, Saskatchewan, and Manitoba)
prior to fall migration. We used generalized linear models to examine infection or evidence
of exposure in relation to host (age, sex, body condition, exposure to other infections),
spatiotemporal (year, province), population-level (local population densities of BWTE, total
waterfowl densities), and environmental (local pond densities) factors. The probability of AIV
infection in BWTE was associated with host factors (e.g., age and antibody status), populationlevel
factors (e.g., local BWTE population density), and year. An interaction between age and AIV
antibody status showed that hatch year birds with antibodies to AIV were more likely to be
infected, suggesting an antibody response to an active infection. Infection with AIV was positively
associated with local BWTE density, supporting the hypothesis of density-dependent
transmission. The presence of antibodies to WNV and APMV-1 was positively associated with age
and varied among years. Furthermore, the probability of being WNV antibody positive was
positively associated with pond density rather than host population density, likely because ponds
provide suitable breeding habitat for mosquitoes, the primary vectors for transmission. Our
findings highlight the importance of spatiotemporal, environmental, and host factors at the
individual and population levels, all of which may influence dynamics of these and other viruses in
wild waterfowl populations.