We tested cloacal and oropharyngeal swab samples from ducks sampled in western Tennessee, United States, during November 7, 2021–January 31, 2022 for IAV infection before and after detection of HP H5N1 at the study site on January 24, 2022. A telemetry study conducted on a subsample of those birds showed no effect of HP H5N1 infection on survival or movement of mallard ducks (13). We collected serum samples (not included in the telemetry study) that enabled us to estimate prevalence of antibodies to IAV nucleoprotein (NP), to IAV HA subtype H5, and to NA subtype N1 before identification of HP H5N1 introduction into the population; document short-term antibody responses in the population after HP H5N1 introduction; examine antibody levels in individual ducks infected with HP H5N1; and address the need to consider population immunity when evaluating effects of HP H5N1 infection on wild avian species. Duck capture and handling procedures in this study were conducted in accordance with Tennessee Technological University’s Institutional Animal Care and Use Committee (IACUC protocol no. 19-20-002) and authorized under Federal Banding Permit nos. 05796 and 24239.

We tested combined cloacal and oropharyngeal swab samples for IAV, including H5 clade 2.3.4.4b, by virus isolation in embryonating chicken eggs and by real-time reverse transcription PCR, as previously described (13). We submitted all H5 nonnegative isolates to the US Department of Agriculture National Veterinary Services Laboratory (Ames, Iowa) for confirmation of subtype and pathogenicity. Duck species tested included mallard (n = 236), northern pintail (A. acuta, n = 19), gadwall (Mareca strepera, n = 4), and American wigeon (M. americana, n = 3). We did not detect HP H5N1 or LP IAV in any samples collected before January 24, 2022 (n = 220). During January 24–31, 2022, the laboratory detected and confirmed HP H5N1 in samples collected from 12/38 (32%) mallards, 3/4 (75%) gadwalls, and 2/2 (100%) American wigeon. Phylogenetic analyses of sequences from a subset of isolated IAV indicated that sequences belonged to genotype A1 of clade 2.3.4.4b viruses, the same genotype as the virus clade first detected in North America in Canada (3).

We tested serum samples for antibodies to IAV NP using a commercial bELISA (IDEXX AI MultiS-Screen AB test; IDEXX Laboratories, https://www.idexx.com). We further tested samples testing positive (sample to negative absorbance ratio <0.7) for antibodies againts H5 and N1 by using hemagglutination inhibition (HI), virus neutralization (VN), and enzyme linked lectin assay (ELLA), as described previously (6,14). We further tested only a subset of samples from the pre–H5N1 detection period. We used 2 attenuated viruses produced by reverse genetics as antigens for HI and VN: reverse genetics BWT, containing HA and NA from the LP A/Blue-winged teal/AI12–4150/Texas/2012 (H5N2), and reverse genetics AST IDCDC-RG71A (H5N8), containing a modified HA and NA from HP 2.3.4.4b A/Astrakhan/3212/2020 (H5N8); remaining gene segments from both viruses were from A/Puerto Rico/8/34. For ELLA, we used A/ruddy turnstone/New Jersey/AI13–2948/2013 (H10N1) as an antigen. We considered samples that tested positive for either antigen at a titer of 32 for HI or 20 for VN positive for H5 antibodies. For ELLA we considered a titer of 80 positive.

IAV NP antibodies were prevalent (79%) in ducks at the study site before introduction of HP H5N1 virus; antibody prevalence increased to 100% after H5N1 was detected (Table 1). Antibody prevalence for H5, based on reactivity to reverse genetics BWT or reverse genetics AST, however, was low before HP H5N1 introduction, as determined by both HI (8%) and VN (25%). By the postdetection sample period, January 30–31, 2022, antibody prevalence increased, as determined by HI (27%) and VN (53%) (Table 1). Antibody prevalence to N1 subtype also increased from 40% before HP H5N1 detection to 86% after detection (Table 1). We observed an increase in the percentage of ducks testing seropositive for H5 after HP H5N1 detection by both HI and VN (Table 2). For H5N1-infected ducks, 15/17 (88%) tested seropositive for NP, 1 (6%) tested seropositive for H5 by HI, 2 (12%) were seropositive for H5 by VN, and 12/16 (75%) tested seropositive for N1. Some of those antibody-positive results, especially in postdetection sampling, could have resulted from seroconversion, but still reflect a very low prevalence of antibodies to H5 at time of infection.

Although antibody prevalence to NP was high in ducks before HP H5N1 virus introduction at the site, likely reflecting previous infections with LP IAV, the prevalence of antibodies to H5 and N1 subtypes was low; few (11 [17%]) ducks tested seropositive for both (Table 1). The lack of immunity to H5 and N1 possibly enabled the successful introduction of HP H5N1 into this population and into North America. That pattern was also apparent in individual ducks that were infected with HP H5N1 at the time we collected serum samples. We detected an increase in antibody prevalence with all serologic tests after HP H5N1 introduction; 62% of birds sampled in the postdetection sample period tested seropositive for both H5 and N1. Whether the developing immunity affected subsequent HP H5N1 transmission as birds migrated from this site is unknown, but many (38%) departing birds apparently remained unexposed. The population was not immunologically naive to IAV at the time of HP H5N1 introduction and could have been partially protected from clinical effects caused by existing heterosubtypic immunity; however, that did not prevent infection or rapid transmission through the population. After HP H5N1 introduction, substantial immunity to H5 and N1 was present. Future studies should address whether existing heterosubtypic or developing homosubtypic immunity affected survival or migration of this duck population; such studies could apply to other wild bird species that may be more vulnerable than ducks to HP H5N1. Monitoring IAV in wild migratory populations can inform influenza spillover risks to domestic animals and humans.