Year‑Round Influenza A Virus Surveillance in Mallards (Anas platyrhynchos) Reveals Genetic Persistence During the Under‑Sampled Spring Season

Abstract

Active influenza A virus (IAV) surveillance in wild waterfowl in the United States has revolved around convenience-based sampling methods, resulting in gaps in surveillance during the spring season. We conducted active IAV surveillance in mallards continuously from July 2017 to July 2019 in the coastal marshes of Lake Erie near Port Clinton, Ohio. We aimed to understand ecological and evolutionary dynamics of IAV across multiple seasons, including the under‑sampled spring season. We collected 2096 cloacal swabs and estimated a 6.1% (95% confidence interval (CI): 0.050-0.071) prevalence during the study period. Prevalence was lowest during spring (1.0%, 95% CI: 0.004-0.015). Time‑stamped phylogenetic analyses revealed local persistence of genetic lineages of multiple gene segments. The PA segment consists of a lineage detected in multiple seasons with a time to most recent common ancestor of 2.48 years (95% highest posterior density: 2.16-2.74). Analysis of the H3 and H6 segments showed close relation between IAVs detected in spring and the following autumn migration. Though the mechanisms behind viral persistence in a single location are not well understood, we provide evidence that viruses can persist across several seasons. Current surveillance methods should be evaluated to ensure they are capturing the breadth of genetic diversity of IAV in waterfowl and prepare for IAV outbreaks in both animals and humans.

Keywords: Anas platyrhynchos; influenza A virus; mallards; phylogenetics; spring migration; surveillance.

Figure 1

Number of samples collected for influenza A virus (IAV) surveillance in mallards and estimated population size of all waterfowl species near Port Clinton, Ohio, USA by month. Year-round IAV surveillance was conducted in wild mallards (Anas platyrhynchos) in one location for two years (July 20217–July 2019) to better capture the under-sampled spring migration. Samples for IAV surveillance collected from mallards by active trapping (January–August) and hunter harvest (September–December) are shown by blue bars. Estimated population size of waterfowl at the study location is shown by the red line. Data is shown as starting January 2018 as estimated population size was not available before that time.

Figure 2

Estimated influenza A virus (IAV) prevalence in migratory mallards near Port Clinton, Ohio, USA by season. Year-round, active IAV surveillance in mallards (Anas platyrhynchos) was conducted for two years (July 2017–July 2019) in one location to fill a gap during spring migration and resulted in 2096 cloacal swabs. Viral isolation was attempted on all real-time reverse transcription polymerase chain reaction positive and undetermined samples. Prevalence was estimated by the proportion of viral isolates and the number of cloacal swabs collected during each season expressed as a percentage. Estimated IAV prevalence of mallards at the study location for each season of the study period is shown by blue bars with 95% confidence intervals shown by red error bars. Summer: June–August; autumn: September–November; winter: December–February; spring: March–May.

Figure 3

Historical influenza A virus (IAV) surveillance near Port Clinton, Ohio, USA by year and season. Year-round, Active IAV surveillance in wild mallards (Anas platyrhynchos) at one location was conducted from July 2017–July 2019 to represent the under-sampled spring season. Historical (autumn 2008–spring 2017) surveillance data for the study location is shown in order to demonstrate the gap in spring surveillance prior to the current study. IAV prevalence for each year by season was estimated by the proportion of IAV isolates and the number of cloacal swabs collected and is shown by the blue bars. Red brackets indicate the current study period. Asterisks (*) represent years for which no samples were collected during that season. Summer: June–August; autumn: September–November; winter: December–February; spring: March–May.

Figure 4

Maximum clade credibility tree of the PA gene segment. Active influenza A virus surveillance was conducted year-round in wild mallards (Anas platyrhynchos) in one location over two years. Time-stamped phylogenetic analysis with a general time reversible plus gamma substitution model revealed genetic persistence across multiple seasons, including the historically under-sampled spring season, in the relatively conserved PA gene segment. Highly supported clades (>0.95 posterior probability) containing viruses detected in the spring (green) who share a common ancestor with viruses detected in another season (red) throughout the study period with a time to most recent common ancestor of <2.5 years are highlighted. Node ages are indicated. Scale bar represents time in years.

Figure 5

Maximum clade credibility tree of the M gene segment. Year-round, active influenza A virus surveillance was conducted in wild mallards (Anas platyrhynchos) in one location over two years. Time-stamped phylogenetic analysis with a general time reversible plus gamma substitution model revealed genetic persistence across multiple seasons, including the historically under-sampled spring season, in the relatively conserved M gene segment. Highly supported clades (>0.95 posterior probability) containing viruses detected in the spring (green) who share a common ancestor with viruses detected in another season (red) throughout the study period with a time to most recent common ancestor of <2.5 years are highlighted. Node ages are indicated. Scale bar represents time in years.

Figure 6

Maximum clade credibility tree of the H3 gene segment. Year-round, active influenza A virus surveillance was conducted in wild mallards (Anas platyrhynchos) in one location over two years. Time-stamped phylogenetic analysis with a general time reversible plus gamma substitution model revealed genetic persistence across multiple seasons, including the historically under-sampled spring season, in the HA gene segment. Highly supported clades (>0.95 posterior probability) containing viruses detected in the spring (green) who share a common ancestor with viruses detected in another season (red) throughout the study period with a time to most recent common ancestor of <2.5 years are highlighted. Node ages are indicated. Scale bar represents time in years.

Figure 7

Maximum clade credibility tree of the H6 gene segment. Year-round, active influenza A virus surveillance was conducted in wild mallards (Anas platyrhynchos) in one location over two years. Time-stamped phylogenetic analysis with a general time reversible plus gamma substitution model revealed genetic persistence across multiple seasons, including the historically under-sampled spring season, in the HA segment. Highly supported clades (>0.95 posterior probability) containing viruses detected in the spring (green) who share a common ancestor with viruses detected in another season (red) throughout the study period with a time to most recent common ancestor of <2.5 years are highlighted. Node ages are indicated. Scale bar represents time in years.