Assessing avian influenza surveillance intensity in wild birds using a One Health lens

Abstract

Wildlife disease surveillance, particularly for pathogens with zoonotic potential such as Highly Pathogenic Avian Influenza Virus (HPAIV), is critical to facilitate situational awareness, inform risk, and guide communication and response efforts within a One Health framework. This study evaluates the intensity of avian influenza virus (AIV) surveillance in Ontario's wild bird population following the 2021 H5N1 incursion into Canada. Analyzing 2562 samples collected between November 1, 2021, and October 31, 2022, in Ontario, Canada, we identify spatial variations in surveillance intensity relative to human population density, poultry facility density, and wild mallard abundance. Using the spatial scan statistic, we pinpoint areas where public engagement, collaborations with Indigenous and non-Indigenous hunter/harvesters, and working with poultry producers, could augment Ontario's AIV wild bird surveillance program. Enhanced surveillance at these human-domestic animal-wildlife interfaces is a crucial element of a One Health approach to AIV surveillance. Ongoing assessment of our wild bird surveillance programs is essential for strategic planning and will allow us to refine approaches and generate results that continue to support the program's overarching objective of safeguarding the health of people, animals, and ecosystems.

Keywords: Avian influenza virus; Migratory birds; One health; Poultry; Wild birds; Wildlife disease surveillance; Zoonotic pathogen surveillance.

Fig. 1

(A) Choropleth map of human population density (number of people per square kilometre, categorized into quantiles) across 577 census subdivisions (CSD). Bivariate choropleth maps of human population density and B) live and harvest and C) sick and dead wild bird sample density (number of wild bird samples collected between November 2021–November 2022 per square kilometre, categorized into quantiles). Locations of spatial clusters of low (blue) and high (red) wild bird surveillance, based on a purely spatial discrete Poisson probability model, shown for D) live and harvest and E) sick and dead wild bird surveillance (number of wild bird samples submitted per CSD between November 2021–November 2022). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

(A) Choropleth map of poultry and egg facility density (number of poultry and egg facilities per square kilometre, categorized into quantiles) across 273 census consolidated subdivisions (CCS) in Ontario, Canada. Bivariate choropleth maps of poultry facility density and B) live and harvest and C) sick and dead wild bird sample density (number of wild bird samples collected between November 2021–November 2022 per square kilometre, categorized into quantiles). Locations of spatial clusters of low (blue) and high (red) wild bird surveillance, based on a purely spatial discrete Poisson probability model, shown for D) live and harvest and E) sick and dead wild bird surveillance (number of wild bird samples submitted per CSD between November 2021–November 2022). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

(A) Choropleth maps of pre-breeding (March–May; Panel 1) and post-fledging (September–November; Panel 2) seasonal abundance for wild mallards in Ontario, Canada using EBird 2021 predictions (Fink et al., 2022), categorized into quantiles. Bivariate choropleth maps of wild mallard seasonal abundance and B) number of live and harvest wild bird samples and C) number of sick and dead wild bird samples submitted over the relevant time period in 2022. Locations of spatial clusters of low (blue) and high (red) wild bird surveillance, based on a purely spatial discrete Poisson model, shown for D) live and harvest and E) sick and dead wild bird surveillance conducted over the relevant time-period. Abundance was not available for areas within the Ontario boundary shown in dark grey. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

(A) Choropleth maps of pre-breeding (March–May; Panel 1) and post-fledging (September–November; Panel 2) seasonal abundance for wild mallards in Ontario, Canada using EBird 2021 predictions (Fink et al., 2022), categorized into quantiles. Bivariate choropleth maps of wild mallard seasonal abundance and B) number of live and harvest wild bird samples and C) number of sick and dead wild bird samples submitted over the relevant time period in 2022. Locations of spatial clusters of low (blue) and high (red) wild bird surveillance, based on a purely spatial discrete Poisson model, shown for D) live and harvest and E) sick and dead wild bird surveillance conducted over the relevant time-period. Abundance was not available for areas within the Ontario boundary shown in dark grey. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Choropleth maps of human population density (number of people per square kilometre, categorized into quantiles) across 577 census subdivisions (CSD) in Ontario, Canada and A) highly pathogenic avian influenza virus (HPAIV) infected domestic premises, B) live and harvest and C) sick and dead wild bird surveillance data collected between November 2021–November 2022.

Fig. 5

Choropleth maps of poultry and egg facility density (number of poultry and egg facilities per square kilometre, categorized into quantiles) across 273 census consolidated subdivisions (CCS) in Ontario, Canada and A) highly pathogenic avian influenza virus (HPAIV) infected domestic premises, B) live and harvest and C) sick and dead wild bird surveillance data collected between November 2021–November 2022.

Fig. 6

(A) Choropleth maps of pre-breeding (March–May; Panel 1) and post-fledging (September–November; Panel 2) seasonal abundance for wild mallards in Ontario, Canada using EBird 2021 predictions (Fink et al., 2022), categorized into quantiles with (B) highly pathogenic avian influenza virus (HPAIV) infected domestic premises, (C) live and harvest and (D) sick and dead wild bird surveillance data collected between March and May 2022. Predicted abundance was not available for areas within the Ontario boundary that are white.

Fig. 6

(A) Choropleth maps of pre-breeding (March–May; Panel 1) and post-fledging (September–November; Panel 2) seasonal abundance for wild mallards in Ontario, Canada using EBird 2021 predictions (Fink et al., 2022), categorized into quantiles with (B) highly pathogenic avian influenza virus (HPAIV) infected domestic premises, (C) live and harvest and (D) sick and dead wild bird surveillance data collected between March and May 2022. Predicted abundance was not available for areas within the Ontario boundary that are white.

Supplemental Fig. 1

Choropleth maps of Ontario, Canada for A) commercial poultry and egg facility density, B) turkey facility density, and C) all other poultry facility density (number facilities per square kilometre) across 273 census consolidated subdivisions (CCS). Densities are categorized into quantiles.

Supplemental Fig. 2

Choropleth maps of overwinter (December – February; Panel 1) and brood rearing (June – August; Panel 2) seasonal abundance for wild mallards in Ontario, Canada using EBird 2021 predictions (Fink et al., 2022) categorized into quantiles with (A) live and harvest and (B) sick and dead wild bird surveillance data collected over the relevant time periods. There were no HPAIV infected domestic premises during these time periods.