Discussion
This paper describes the global AIV subtype diversity and distribution in water. It was evidenced that wild bird habitats host the highest subtype diversity reported in water samples. Wild birds of aquatic environments such as the Anseriformes and Charadriiformes constitute the major natural AIV reservoir. At least 105 wild bird species of 26 different families host LPAI, and almost all the AIV subtypes have been detected in wild aquatic bird reservoirs (Olsen et al., 2006). Wild waterfowl greatly contribute to the geographical spread of AIV subtypes between wetlands through migratory movements associated with high bird densities and increased contact rates among bird species (van Dijk et al., 2018).
Globally, a total of 112 AIV subtypes have been identified in wild birds of which 49 have been found in domestic birds (Olson et al., 2014). Only in China, twelve subtypes of HA (H1-H12) and eight subtypes of NA (N4-N9) of 21 different combinations have been identified in wild birds (Tang et al., 2020). In our analysis, we evidenced that 21 subtypes were reported from water samples, of which eleven were detected in wild bird habitats, mainly from Asian countries. Large parts of East Asia and all Southeast Asia comprise the East Asian-Australasian Flyway. This flyway supports the greatest diversity and highest number of migratory birds worldwide (Tang et al., 2020). Likewise, a large number of wild aquatic birds with a great potential to serve as carriers of AIV migrate to overwintering habitats in Asia annually (Deng et al., 2013; Khalil et al., 2020; Nakagawa et al., 2018). Nevertheless, it is well-known that geographical variations on existing surveillance efforts can also potentially influence the AIV detection and its spatial distribution (Berger et al., 2018).
The Izumi plain and the Dongting Lake are recognized as prominent overwintering sites in Japan and China, respectively (Nakagawa et al., 2018; Zhang et al., 2011). These stopover areas for several tens of thousands of migratory birds represented the wild bird habitats with the highest number of HA sequences and the greatest subtype diversity, mainly the Izumi plain; wherein, six different subtypes (H3N8, H4N6, H4N8, H5N6, H5N8, and H6N2) were reported in water samples. Moreover, in our phylogenetic tree, most of the viruses from the Izumi plain were related to viruses from poultry farms of the Dongting Lake region in China and from turkey barns in the United States, as well as from Chinese live bird markets. Previous phylogenetic analyses of AIV strains isolated from the Izumi plain have revealed genetic reassortments between AIV from East Asian, European, African, and North American countries (Khalil et al., 2020, 2021; Nakagawa et al., 2018; Okuya et al., 2015). This trans-hemispheric genetic flow of AIV highlights the wild-domestic bird interfaces as relevant areas for influenza A virus surveillance (Prosser et al., 2013).
The Izumi plain and the Dongting Lake are also well-known for free-range farming and mixing between chickens and domestic aquatic fowl (Khalil et al., 2020). In our analyses, all the HA sequences from the subtypes H3N2, H3N8, H4N9, H11N2, and H12N7 isolated in water samples from poultry farms in China were close to aquatic environments. Domestic farming in Southern Asia commonly occurs with a lack of biosecurity measures and close contact among wild waterfowls and domestic fowls that facilitates multiple genetic reassortments (Deng et al., 2013; Zhang et al., 2011). This type of farming is widespread in low-income countries where the majority of poultry is raised under extensive conditions by family-based smallholder farms (Gilbert et al., 2015). Therefore, regions such as Eastern Europe, Central America, and sub-Saharan Africa also pose a high potential for AIV interspecies transmission; however, several high-risk areas have inadequate influenza A virus surveillance (Berger et al., 2018).
Likewise, live bird markets where different domestic and wild bird species often share the same water, food, and housing also represent an opportunity for interspecies transmission and viral genetic diversification (Zhang et al., 2011). In our analyses, most of the HA sequences were detected in water samples from live bird markets, mainly from H9N2 and H5N1 subtypes in Asia (China and Bangladesh). For more than a decade, Asian countries have undertaken numerous efforts to rapidly detect and track AIVs mainly for LPAI H9N2 and HPAI H5N1 through annual surveillance programs in poultry-related environments such as live poultry markets, poultry farms, slaughterhouses, and wild bird habitats (Rimi et al., 2019; Zhang et al., 2019). These efforts provide valuable information to inform decision-making and implement risk mitigation strategies. Unfortunately, other countries and regions do not have the same level of surveillance or do not share that data publicly, which hampers the possibility to better understand AIV transmission dynamics locally and globally (Chan et al., 2010).
The H5N1 virus continues to pose an important public health threat in East and Southeast Asian countries and has become endemic in domestic poultry in these countries (Rimi et al., 2019). Although HPAI H5N1 has also caused many outbreaks with severe illness in poultry from Europe and Africa (Chowdhury et al., 2019), in our analyses HPAI H5N1 viruses were only reported in water samples from Asian countries. Similarly, in our study LPAI H9N2 viruses were isolated solely in water samples from Asia. Nevertheless, LPAI H9N2 avian influenza viruses have widespread in domestic poultry worldwide (Peacock et al., 2019).
Multiple AIV subtypes have been detected from poultry and wild birds in Africa, Australia, Europe, and America (Alexander, 2007; Araujo et al., 2018; Brown, 2010; Grillo et al., 2015; Hansbro et al., 2010; Jiménez-Bluhm et al., 2018; Senne, 2007). However, our global analysis of AIV evidenced a lack of sequences isolated from environmental water in Africa, Oceania, and South America, as well as, in Europe. Data on influenza in tropical countries remain scarce compared with temperate countries (Moura, 2010). Likewise, according to Moura (2010), the emergence of AIV strains in Asia may occur and be detected approximately 6 to 9 months earlier than in Oceania, North America, and Europe, and 12 to 18 months earlier than in South America.
It is also recognized that the prevalence and distribution of avian influenza in most tropical countries are mostly unknown as a result of the lack of a rigorous surveillance system (Yazdanbakhsh & Kremsner, 2009). Since the H1N1 pandemic, it has been highlighted a considerable number of shortcomings on global epidemiological surveillance. The absence of routine AIV monitoring has resulted in substantial information gaps in large areas of the world, mainly from less-resourced countries (Briand et al., 2011).
Environmental sampling has been effectively used for AIV surveillance since the 1970s. Nevertheless, the methods and protocols are not completely standardized, as well as, an international guideline about data management is absent (Hood et al., 2020). This was one of the most noticeable limitations in our analyses since several AIV no specified the type of environment sampled (i.e., feces, water, air, mud, or surface swabs), and thus, those HA sequences were excluded. Likewise, we only analyzed the sequences with complete subtypes reported in the four recognized electronic databases. In the case of Europe, there were reported 83 sequences of H1 genes from ice and water in high-latitude lakes visited by large numbers of migratory birds in Siberia. Nevertheless, those sequences did not meet the inclusion criteria. We also found several geographical regions with missing subtype data. Therefore, the AIV subtypes described in this work may not be the exact reflection of the global subtype diversity, but it highlights the potential value of using this information to better understand AIV local and global transmission dynamics. Our work also highlights the need to improve surveillance efforts in many regions as well as to advance towards more unified data collection and sharing standards to improve influenza A virus surveillance and better prevent future potential pandemics.
In conclusion, this descriptive and phylogenetic analysis of AIVs isolated in water samples from sites at high risk for influenza outbreaks, such as live bird markets, poultry farms, and wild bird habitats is valuable to provide an overview and baseline information of the current data on global AIV diversity and distribution since 2003. However, this study highlights the need to continuing generating, expanding, and sharing precise and detailed environmental data from surveillance systems to allow a better understanding of the ecology and epidemiology of AIV, especially from low- and middle-income countries.