Avian influenza viruses in New Zealand wild birds, with an emphasis on subtypes H5 and H7: Their distinctive epidemiology and genomic properties

Abstract

The rapid spread of highly pathogenic avian influenza (HPAI) A (H5N1) viruses in Southeast Asia in 2004 prompted the New Zealand Ministry for Primary Industries to expand its avian influenza surveillance in wild birds. A total of 18,693 birds were sampled between 2004 and 2020, including migratory shorebirds (in 2004-2009), other coastal species (in 2009-2010), and resident waterfowl (in 2004-2020). No avian influenza viruses (AIVs) were isolated from cloacal or oropharyngeal samples from migratory shorebirds or resident coastal species. Two samples from red knots (Calidris canutus) tested positive by influenza A RT-qPCR, but virus could not be isolated and no further characterization could be undertaken. In contrast, 6179 samples from 15,740 mallards (Anas platyrhynchos) tested positive by influenza A RT-qPCR. Of these, 344 were positive for H5 and 51 for H7. All H5 and H7 viruses detected were of low pathogenicity confirmed by a lack of multiple basic amino acids at the hemagglutinin (HA) cleavage site. Twenty H5 viruses (six different neuraminidase [NA] subtypes) and 10 H7 viruses (two different NA subtypes) were propagated and characterized genetically. From H5- or H7-negative samples that tested positive by influenza A RT-qPCR, 326 AIVs were isolated, representing 41 HA/NA combinations. The most frequently isolated subtypes were H4N6, H3N8, H3N2, and H10N3. Multivariable logistic regression analysis of the relations between the location and year of sampling, and presence of AIV in individual waterfowl showed that the AIV risk at a given location varied from year to year. The H5 and H7 isolates both formed monophyletic HA groups. The H5 viruses were most closely related to North American lineages, whereas the H7 viruses formed a sister cluster relationship with wild bird viruses of the Eurasian and Australian lineages. Bayesian analysis indicates that the H5 and H7 viruses have circulated in resident mallards in New Zealand for some time. Correspondingly, we found limited evidence of influenza viruses in the major migratory bird populations visiting New Zealand. Findings suggest a low probability of introduction of HPAI viruses via long-distance bird migration and a unique epidemiology of AIV in New Zealand.

Fig 1. Map of migration routes of waders to and from New Zealand (Credit: Adrian Riegan, Pūkorokoro Miranda Naturalists’ Trust).

Fig 2. New Zealand sites where wild birds were captured for sample collection for avian influenza surveillance testing, 2004–2020.

Fig 3. Prevalence of H5 and H7 AIVs as determined by RT-qPCR in mallard ducks in New Zealand, 2004–2020.

Fig 4. Prevalence of influenza A, H5, and H7 genes as detected by RT-qPCR in mallard ducks in New Zealand, 2012–2020.

Fig 5. Summary tree for the H5 subtype derived from Bayesian analysis using a HKY substitution model, a strict clock, and a constant coalescent model.

Maximum clade credibility was used to construct the summary tree. The sequences are coloured according to the sampling location, i.e., Asia (yellow), Africa (red), Europe (dark green), North America (purple), South America (pink), Australia (lime green), or New Zealand (blue). Sequences with high similarity from the same country are down-sampled for readability.

Fig 6. Timed phylogeny of the summary tree for the H5 subtype, using an HKY substitution model, a strict clock, and a constant coalescent model.

The 95% highest posterior density estimates of the most recent common ancestors (MRCAs) for the New Zealand clade, New Zealand’s sister clade (the North American clade), and the Australian clade are shown as bars. The time scale is in years. Sequences with high similarity from the same country are down-sampled, and the most distant clade (the Eurasian/African/North American clade) is collapsed for improved readability.

Fig 7. Summary tree for the H7 subtype from Bayesian analysis using an HKY substitution model, a strict clock, and a constant coalescent model.

Maximum clade credibility was used to construct the summary tree. The sequences are colored by sampling location, i.e., Asia (yellow), Africa (red), Europe (dark green), North America (purple), South America (pink), Australia (lime green), or New Zealand (blue). Sequences with high similarity from the same country are down-sampled for readability.

Fig 8. Timed phylogeny of the summary tree for the H7 subtype HA gene, using an HKY substitution model, a strict clock, and a constant coalescent model.

The 95% highest posterior density estimates of the most recent common ancestors for the New Zealand clade, New Zealand’s sister clade (the Eurasian clade), and the Australian clade are shown as bars. The time scale is in years. Sequences with high similarity from the same country are down-sampled for readability.