Dept title


H5N1 HPAI in Different Species

This chapter reviews the role of various species in the genesis, persistence and spread of Asian-lineage H5N1 HPAI viruses. Special attention is paid to the domestic duck (Anas platyrhynchos var. domesticus), which is considered to have played a key role in these processes. An understanding of the role of domestic ducks in the maintenance and spread of these viruses helps to explain why the current outbreak of HPAI is panzootic whereas earlier outbreaks were not.

3.1 H5N1 HPAI and domestic ducks

Ducks (wild and domestic) are capable of being infected subclinically with a wide range of avian influenza viruses but until recently most isolates from ducks were of low pathogenicity (see for example Stallknecht et al, 1990b).

Prior to the emergence of Asian-lineage H5N1 HPAI viruses there were few reports of natural infection with HPAI viruses in ducks. One noteworthy exception occurred in Ireland in 1984, involving an HPAI virus of the H5N8 subtype. This highly pathogenic virus was detected in clinically normal domestic ducks on a farm adjacent to a turkey farm that several months earlier had experienced outbreaks of HPAI (Alexander et al, 1987).

Experimental infections of ducks with HPAI viruses (other than Asian-lineage H5N1 HPAI viruses) have yielded mixed results. In one study, ducks were relatively resistant to experimental challenge to a range of HPAI viruses (Alexander et al, 1978), although other studies found that ducks could be infected subclinically. Despite the lack of clinical signs, some of these infections were systemic (see for example Kawaoka et al, 1987; Nestorowicz et al, 1987). These experimental studies and the earlier field experience in Ireland demonstrated the potential for clinically normal but infected ducks to propagate and spread HPAI viruses.

Studies conducted in Hong Kong prior to 1997 did not lead to detection of avian influenza viruses of the H5N1 subtype in domestic ducks, suggesting that the emergence of Asian-lineage H5N1 HPAI viruses is a relatively recent event. Over a five-year period, between 1975 and 1980, samples were collected from domestic ducks (8 737), geese (1 353), and chickens (1 708) originating from southern China and Hong Kong. A total of 586 avian influenza viral isolates were obtained. Of these, 22 were H5N3 subtype (21 from ducks and one from a goose), and one (from a duck) was an H5N2 subtype (Shortridge, 1992). These H5 viruses were not direct precursors of the Asian-lineage H5N1 HPAI viruses.

The first recorded cases of infection with H5N1 HPAI viruses in domestic ducks in Asia were in 1997 from live-bird markets in Hong Kong SAR (Shortridge, 1999). These positive samples were collected in late December 1997, at a time when levels of infection in terrestrial poultry in markets were extremely high. It is possible that these cases resulted from spillover of infection from terrestrial poultry back to ducks, rather than ducks being the initial source of the virus for the chickens (the percentage of chicken samples yielding virus was far greater than the percentage of positive samples from ducks). This was supported by subsequent experimental infection with a 1997 H5N1 virus isolate from Hong Kong SAR (HK/97 strains) which indicated that ducks were relatively resistant to infection with HK/97 strains of virus, although virus was recovered from oropharyngeal swabs and from the lung of one inoculated duck (Shortridge et al, 1998; Perkins and Swayne, 2002). This suggested ducks played a minimal role in the perpetuation of the 1997 Hong Kong SAR H5N1 HPAI viruses. Once these viruses were eliminated from Hong Kong SAR in 1997, no H5N1 viruses with the same combination of internal genes and NA gene (HK/97-like viruses) were detected in imported ducks (or indeed other poultry), despite regular intensive surveillance.

The first case of infection with an H5N1 HPAI virus in domestic ducks in mainland China was not reported until 1999 (Chen et al, 2004). This sample was reputedly from a healthy duck on a farm in Guangxi Autonomous Region. Seven of the eight genes of this virus were Goose/Guangdong/96-like but the virus had acquired a different NS gene from an unknown source.

By late 2000, healthy ducks transported to Hong Kong SAR from mainland China were found to be infected with H5N1 HPAI viruses (Guan et al, 2002a). These viruses differed from those seen previously in that they were reassortants, having acquired internal genes from other influenza viruses, presumably from aquatic birds. This was the first indication that apparently healthy ducks excreting H5N1 HPAI viruses could spread infection across territorial boundaries.

The first reported case of infection with a so-called ‘Z’ genotype H5N1 virus was apparently from a normal farmed duck in 2001, also in Guangxi Autonomous Region (Sims et al, 2005). In an earlier publication (Chen et al, 2004) this was referred to as a ‘G’ genotype virus.

In 2001, an H5N1 virus was detected in meat from a duck imported to the Republic of Korea from China (Tumpey et al, 2002). This case again demonstrated that ducks healthy enough to be presented for slaughter could be systemically infected, and also showed the potential for long-distance transport of virus through trade in poultry meat. Subclinical systemic infection with H5N1 HPAI viruses was confirmed through experimental inoculation of susceptible ducklings (Tumpey et al, 2002). The virus was capable of replication in many tissues of inoculated ducks, including muscle and brain. Compared with some earlier H5N1 HPAI viruses isolated from poultry it multiplied to higher titres in lung and kidney.

Fatal disease in ducks caused by H5N1 HPAI viruses was not reported until 2002-03 and involved Anatidae (and other orders) in two zoological collections in Hong Kong SAR. This involved two separate genotypes (Z and Z+) suggesting different origins for these two outbreaks, which were located some 12 km apart (Ellis et al, 2004; Sturm-Ramirez et al, 2004).

Studies in Thailand in 2004 demonstrated that, under field conditions, not all infected ducks developed signs of disease (Songserm et al, 2006). Experiments conducted elsewhere demonstrated that the outcome of inoculation depended on the strain of virus (Sturm-Ramirez et al, 2005) although specific molecular changes associated with the increased capacity to produce disease in ducks were not determined. Others have demonstrated an age-related difference in susceptibility (Swayne and Pantin-Jackwood, 2006; Pantin-Jackwood and Swayne, 2007).

In ten flocks of grazing ducks in Thailand that were examined in detail, infection was not detected while ducklings were being brooded but occurred within 12 to 63 days after the birds were released for grazing. It was not determined how these ducks became infected (Songserm et al, 2006).

The detection of infection in subclinically infected ducks in Thailand was achieved through intensive targeted surveillance. Enhanced surveillance starting in October 2004 included apparently healthy ducks and led to a marked increase in the number of cases detected (Tiensin et al, 2005).

Ducks are now known to be capable of excreting H5N1 HPAI virus via the cloacal and respiratory routes for at least 17 days (Hulse-Post et al, 2005). During this time, if they contact other poultry either directly or indirectly, they could potentially spread infection, assuming the quantity of virus excreted exceeds the infectious dose for the exposed poultry.

Recent studies have demonstrated the presence of virus in feather epithelium of call ducks experimentally infected with H5N1 HPAI virus, suggesting that shed feathers could pose a potential risk for spread of infection (Yamamoto et al, 2007).

3.1.1 Other relevant information on management and marketing of ducks

The numbers of ducks reared in China increased more than three-fold from a standing population of 223 million in 1985 to 725 million in 2005 (FAO, 2006) with many of these reared on ponds, potentially allowing contact with wild birds. This increase in the number of domestic ducks increased the number of susceptible poultry in which H5N1 viruses could multiply and, given the predominance of live-bird marketing and lack of segregation of species, probably increased the chance of spread of these viruses to other poultry and humans. Recently, this susceptibility has been reduced to some extent through the use of vaccination and, in some places, through segregation of terrestrial and aquatic poultry in markets and during transportation.

The number of ducks in Viet Nam also increased in the latter part of the 20th century and early 21st century (until numbers fell in 2004-05 as a result of HPAI control measures, including breeding bans) but not to the extent seen in China (approximately four percent annual growth over the previous nine years) (FAO, 2006) . Most of these ducks are reared on channels, ponds and paddy fields with very few reared on enclosed intensive farms.

China and Viet Nam account for 75 percent of the world’s duck population (about 775 million of the approximate standing population of 1.044 billion (FAO 2006). Both Indonesia (around 34 million) and Thailand (around 17 million) also have relatively large duck populations. Overall, approximately 90 percent of the world’s domestic ducks are in Asia (computed from FAO, 2006).

Ducks are relatively high value animals and are transported over long distances to markets. For example, in China it is known that they travel more than 400 km from inland provinces such as Hunan to coastal markets in Guangdong (see for example Li et al, 2004a).

Grazing of ducks on paddy fields has been practised for many years but the development of a specific grazing industry involving the long-distance trucking of ducks across Thailand is a relatively recent phenomenon. Until about 30 years ago, this was apparently a local practice with limited movement of ducks across provincial boundaries (D. Hoffman, personal communication).

Most ducks in Asia are reared extensively and this is considered to be a key factor in the transmission of H5N1 HPAI viruses. In one study in Thailand of infection in ducks reared using different management systems, there were no reports of disease or infection in ducks reared intensively indoors despite sampling of all batches of ducks prior to market (60 samples for virus detection per flock). By contrast, for ducks reared in open sheds, four of 17 tested flocks were infected, as were 28 of 61 grazing flocks (Songserm et al, 2006).

In intensive (so called 'X-ray') surveillance of village flocks in Thailand conducted in October 2004, 47 percent of backyard duck flocks were found to be ‘infected’ and were culled (Songserm et al, 2006). This demonstrated the high level of exposure of these ducks to infection at that time and suggests they played a significant role in the maintenance and spread of H5N1 HPAI viruses.

In Thailand, levels of infection in ducks appear to have been reduced through restrictions on grazing and culling of infected flocks (Tiensin et al, 2007a). These measures correlated with a reduction in cases in other poultry suggesting, but not proving, that ducks were playing an important role in the transmission of these viruses to other species.

3.1.2 Geospatial match between disease and ducks

An apparent association between ducks and outbreaks of disease has also been demonstrated through geospatial mapping studies (Fig 1).

Several studies have demonstrated an association between areas where ducks are allowed to graze (i.e. areas where rice cropping is practised and especially those areas with multiple rice crops every year) and the occurrence of H5N1 HPAI (e.g. Gilbert et al, 2006a). These studies do not prove that domestic ducks are the only factor involved in the spread of the disease, but suggest a strong association. Other confounding factors associated with wetland agriculture could also be involved, notably wild birds which share the same ecosystem and, potentially, could be short-term subclinical or mechanical carriers of virus.

A similar occurrence is found in Viet Nam where most of the domestic ducks are located in the Mekong River delta and the Red River delta – the areas most affected by H5N1 HPAI. Here, attempts have been made to reduce the levels of infection in ducks through a combination of vaccination and restrictions on breeding and marketing, although recent outbreaks in ducks (in late 2006 and throughout 2007), especially in unvaccinated ducks, indicate that there is still scope to improve application of these preventive measures.

In China, the majority of ducks are reared in the south of the country, which is purported to be the 'epicentre' for emergence of many avian influenza viruses (Webster and Govorkova, 2006). However, the poultry subsector in the south of China also differs from that in the north with higher numbers of local 'yellow breeds' of chicken, and probably more sales of poultry through live-poultry markets (Sims et al. 2005). Thus, it is possible that these other factors also contributed to the higher levels of infection with H5N1 HPAI viruses that have been recorded in this region.

The information presented above suggests but does not prove a causal link between rearing of ducks outdoors and spread of H5N1 HPAI. However, when combined with experimental and field data showing that individual ducks can be infected 'silently' and excrete low levels of virus for a short period of time (usually about 1-2 weeks), it suggests that domestic ducks have played and perhaps continue to play a crucial role in persistence and spread of infection.

Fig. 1 HPAI, ducks and rice in Southeast Asia

3.1.3 Mixing of ducks and other species

Mixed poultry farming is widespread across Asia especially in the village agricultural sector but also extends to some commercial flocks. Segregation of terrestrial and aquatic poultry is not practised on many small farms, during transport to market or in some live-poultry markets.

To assist in preventing the spread of HPAI in Hong Kong SAR, a complete segregation policy was imposed in 1998 for domestic waterfowl (Sims et al, 2003) and today it is still illegal to rear, transport or market ducks and geese together with other poultry. The value of this segregation was demonstrated between 1999 and 2000 when infection was detected in clinically normal geese and ducks sent from Guangdong province of China for slaughter in Hong Kong SAR but no infection was detected in terrestrial poultry that were marketed and transported separately during the same period. It was not until February 2001 that infection recurred in terrestrial poultry in markets in Hong Kong SAR.

Hence, it is likely that control of H5N1 HPAI in Asia and elsewhere would be facilitated by the separation of domestic waterfowl from terrestrial chickens, at least in markets and commercial farms. The difficulty of achieving this at the village level may prevent implementation in small flocks of scavenging poultry; alternative, practical means of control and prevention of infection may need to be explored in these situations.

3.1.4 Ducks and wild birds

Of all the types of poultry reared, domestic ducks are those most likely to have direct and indirect interactions with wild birds, given their shared habitat in wetlands/paddy fields. This provides opportunities for an exchange of viruses (in both directions) between these two populations. Water birds are not the only wild birds that can contact ducks in paddy fields. In Thailand, for example, many other types of free-flying birds can be found on paddy fields and these could play a potential role in dissemination of virus if they forage in areas frequented by infected ducks, even if just mechanical or short-term carriers of virus over relatively short distances. Recent experimental studies with Asian-lineage H5N1 HPAI viruses have shown that starlings (Sturnus vulgaris) and sparrows (Passer domesticus) can be infected for a short period with these viruses and that not all infected starlings die (Boon et al, 2007).  It is not known whether these birds can carry or excrete sufficient quantities of virus to infect poultry, especially since some species of birds, such as wood ducks and turkeys, appear to require a lower infectious dose than chickens (Brown et al, 2007a, I. Brown, personal communication).

3.1.5 Ducks and water

Infected ducks transfer H5N1 HPAI viruses to the ponds, fields or wetlands they inhabit. Further, these viruses can survive in such environments for a variable length of time determined largely by temperature (three days in paddy field water at 25-32oC [Songserm et al, 2006] and longer under cooler conditions [Brown et al, 2007b]). The salinity of the water and possibly the levels of exposure to UV radiation may also play a role (Brown et al, 2007b, Stallknecht et al, 1990a). Asian-lineage H5N1 HPAI viruses appear to be less well adapted to aquatic conditions than LPAI viruses isolated from aquatic birds (Brown et al, 2007b).

Domestic ducks derived originally from Mallards (e.g. Pekin ducks) are the main species reared commercially around the world and are referred to as "dabblers". Dabblers tend to feed superficially (skimming the surface of water for feed), but can also feed on and filter mud in shallow waterways. This means that virus particles would have to be either suspended in water or deposited in mud in shallow water for these domestic ducks to become infected by this route. Ducks in water are also reputed to practise ‘cloacal sipping’ (in which water is sucked into the cloaca), which could potentially enhance spread of infection if water is contaminated. However, no studies have been conducted on the dispersal of H5N1 virus particles deposited in pond or lake water. The rate of survival of virus in the littoral zone of lakes in places where infection is known to occur is an important area for future study.

The fate of respiratory-borne virus from ducks in water is not known. Since ducks are gregarious animals, the shift towards increased excretion of H5N1 virus via the respiratory route could potentially facilitate duck-to-duck transmission when birds are in close contact. However, studies of rates of transmission between ducks of viruses excreted predominantly via the cloacal or tracheal route have not been conducted.

3.1.6 Ducks and vaccination

Under laboratory conditions, properly formulated vaccines have been found to be successful at reducing and even preventing the shedding of H5N1 viruses in inoculated immune ducks subsequently challenged with virus (see for example Middleton et al, 2007; Beato et al, 2007). Given the many factors that complicate responses to vaccination, it is unlikely that the use of vaccines in the field will lead to elimination of virus, but experimental studies have shown that quantities of virus excreted can be markedly reduced such that transmission of H5N1 in a vaccinated flock would be far less efficient (and may not occur) than in flocks that remain unvaccinated. Vaccination is being used for containment of infection in ducks in China and Viet Nam but no controlled studies have been conducted to determine its overall efficacy under field conditions. The number of reported cases of disease in poultry and humans has fallen since widespread vaccination was introduced in both countries, but it has not been proved that this improvement was due solely to vaccination. In Viet Nam, outbreaks of H5N1 HPAI in 2007 mainly affected unvaccinated ducks.

In discussions on vaccination of ducks, some concerns have been expressed that the use of vaccines will drive antigenic change in H5N1 HPAI viruses. This is possible, but antigenic changes will also be driven by subclinical systemic infection in ducks, which also occurs when vaccination is not used. Selection of variant viruses has been demonstrated in experimental studies in unvaccinated ducks (Hulse-Post et al, 2005).

3.1.7 Duck eggs and day-old ducklings

No studies have been conducted on the presence of Asian-lineage H5N1 HPAI viruses in eggs from infected ducks but, based on experiences with other species, eggs laid by infected ducks could potentially contain some virus (Promkuntod et al, 2006). The surface of duck eggs is frequently soiled with faeces and since ducks can be subclinically infected and pass virus in faeces, contaminated eggs pose a potential transmission risk.

One investigation of cases of disease associated with H5N1 HPAI virus in young ducks in Viet Nam in 2007 suggested an association with specific hatcheries, but this was never demonstrated (see 3.3.2) (D. Hadrill, personal communication). Small-scale duck hatcheries in Viet Nam and elsewhere have the potential to transfer virus to newly-hatched ducklings through the use of straw and recycled transport containers (Sims, unpublished). Trade in these ducklings often involves several ‘middlemen’, who accumulate ducks from multiple sources, increasing the scope for contamination.

No studies have been published on the effect of maternal antibody to H5 avian influenza viruses on infection and virus excretion in day-old ducklings subsequently exposed to an H5N1 HPAI virus, and therefore the extent to which exposed ducklings with maternal antibody carry and excrete virus is not known.

>>> Back to top