Migratory Birds As Scapegoats: The Role of Wild Birds in Spreading Avian Flu

There are several ways in which H5N1 can be spread within and between countries. Three major potential routes are the movements of infected poultry (and poultry products), movements of caged wild birds in trade and movements of wild birds. Effective responses need to focus on all of these possible means of spread.

Outbreaks among wild birds in Europe and Iran during 2006 show that wild birds are capable of carrying the virus to new sites after infection, possibly in a "leap-frog" fashion by travelling for a short time and passing on infection to another group of birds before dying. Many questions remain concerning the effects of the virus on wild birds and the efficiency with which they can spread it to other wild birds or to domestic poultry.

By contrast, recent outbreaks in Cameroon, Egypt, India, Israel, Jordan, Niger, Nigeria and Pakistan originated within the poultry industry. Here, as in most other H5N1 outbreaks, there is strong circumstantial evidence that movements of poultry and poultry products are responsible. The timing and location of these outbreaks do not match the movements of migratory birds. Moreover, in many of these countries, poultry outbreaks occurred almost simultaneously in multiple large-scale poultry operations, indicating that migratory birds were an unlikely agent of the transmission.

For Southeast Asia, recent comprehensive analysis of viral lineages concludes that poultry movements were responsible for multiple reintroductions, both within and between countries, and that transmission within poultry is the major mechanism for sustaining H5N1 in the region.

In some parts of the world, authorities have proposed attempting to control the spread of H5N1 by culling wild birds, or destroying their habitats, or displacing them from breeding and roosting grounds. These approaches are unlikely to prevent the transmission of the disease and may in fact spread it to non-infected areas by forcing already-infected birds to disperse. They are at best ineffective, probably counterproductive, and distract from more suitable interventions. They could also add to the stresses already imposed on some species through habitat loss.

The Role of Wild Birds

In 2006, wild bird outbreaks have occurred across Europe, and sporadic new incidents continue to be reported. Nearly all incidents involve just one or a few individual birds (usually less than 10) apart from the wild bird outbreak on Rügen Island (Germany) which killed over 100 birds.

Outbreaks among wild birds in Europe during 2006 show that wild birds are capable of carrying the virus to new sites after infection - at least during the early stages of infection. It is possible that the initial outbreaks in Europe in February related to movements of birds away from the Black and Caspian Sea regions in response to unusually cold weather. These areas are known to have widespread H5N1 infection in poultry and limited biosecurity measures in place.

In contrast to the recent European incidents, the movement of wild birds is not to date the main cause of the spread of H5N1 in Asia over the decade since the virus was first discovered there. Prior to April 2005, wild birds found dead or dying with H5N1 in Asia were largely sedentary species that scavenge near poultry, live markets or captive bird populations.

From May 2005 to July 2005, there were significant die-offs of migratory wild birds in Qinghai Lake (1,500-6,300 birds) and Mongolia (126 birds). The majority of the birds affected in May 2005 in Qinghai Lake were Bar-headed Geese. Deaths from H5N1 in the geese occurred weeks after their arrival from wintering grounds in India, suggesting that the source of the H5N1 infection was local to Lake Qinghai. Furthermore, no Bar-headed Geese or other wild birds were found dead in other wetlands near to Qinghai Lake.

In Mongolia, at Lake Erhel, the main species found dead or dying with H5N1 in July 2005 were Bar-headed Geese and Whooper Swans. Both species would have arrived to breed in Mongolia several months earlier, and during the outbreak the birds would have been near to completing their annual feather moult, during which they are sedentary. Bird surveys carried out during the same period at eight wetlands in Mongolia found no other birds with H5N1. These facts point to the source of H5N1 infection being local to Lake Erhel and suggest that the infected wild birds did not spread the disease to new locations.

It is possible that in some Asian and Australasian countries where little or no surveillance work has been done, infected wild birds have gone undetected. However, in Japan, Korea, New Zealand and Australia, where there has been extensive surveillance, there have been no infected wild birds identified. Countries such as the Philippines, on a major migration route from Southeast Asia, remain free of H5N1. Japan and South Korea have both remained free of the disease after early outbreaks confined to poultry were brought under control by closing borders to poultry imports.

Wild birds have tested positive for H5N1 in Asia. Six out of more than 13,000 wild birds tested in China were positive, and 3 percent of the 13,000 had antibodies to H5N1. However, out of 16,000 wild birds tested over the last decade in Hong Kong, not a single live bird infected with H5N1 has been found. This finding is particularly striking since Hong Kong is so close to centers of infection in poultry in mainland China.

The lack of a trail of H5N1 infections along migratory pathways from infected breeding habitats in Mongolia, China and Russia to southern wintering areas in Asia suggests that migratory wild birds are not spreading the disease long distances between continents. With few exceptions, there is limited correlation between the pattern and timing of spread among domestic birds and wild bird migrations.

Nevertheless, some authorities argue that the timing and location of outbreaks in the Black Sea and Caspian Sea in the autumn of 2005 did follow the southern migration routes of birds. These outbreaks occurred after H5N1 was detected in poultry in Russia and wildfowl in southwest Russia in the summer of 2005. However, there appeared to be no trails of death along this migration route; wild bird deaths were localized and, in some instances, were restricted to a few individuals from larger flocks. Equally plausible explanations for the spread of avian influenza westwards during the latter half of 2005 are the movements of poultry and poultry products.

It is unlikely that migratory wild birds carried H5N1 to Africa. There is strong circumstantial evidence that movements of poultry and poultry products are responsible. The timing and location of these outbreaks do not match the movements of migratory birds. Moreover, in countries such as Nigeria and Egypt, poultry outbreaks occurred almost simultaneously in multiple large-scale poultry operations, indicating that migratory birds were an unlikely agent of the transmission. If H5N1 was carried by wild birds, outbreaks should have occurred in key wetlands for migratory birds, especially in East Africa where there has been surveillance of wild birds in place over the past six months.

Understanding of the epidemiology of H5N1 in wild birds, and the behavior of the virus in the wider environment, remain very inadequate. Most of the research on H5N1 has been on domestic animals in laboratory environments. How easily infected wild birds can pass the disease on to other wildfowl or poultry remains an important, unanswered question. The limited evidence that exists suggests tremendous variability in transmission rates and virulence between different host species and different strains of the virus.

Better quality data collection and reporting is crucial to understanding general patterns in outbreaks, possible routes of transmission and the potential impacts on migratory bird populations. This information can be used to focus contingency efforts, predict future outbreaks and guide effective policy to reduce the economic and conservation impacts of avian influenza.

The Poultry Trade

Most outbreaks in Southeast Asia can be linked to movements of poultry and poultry products (or accidental transfer of infected material from poultry farms, such as water, straw or soil on vehicles, clothes and shoes). Globally, the most important route of spread remains unrestricted poultry movements. A recent paper presented at the U.S. National Academy of Sciences analyzes the viral lineages and concludes that poultry movements were responsible for multiple reintroductions in Southeast Asia, both within and between countries.

Live animal or "wet" markets may have played a major part in spreading the virus in Southeast Asia, according to the UN Food and Agriculture Organization (FAO), World Organization for Animal Health (OIE) and World Health Organization (WHO).

There is also a huge international trade in poultry - both legal and illegal. The legal trade involves literally millions of hatching eggs and poultry being shipped to destinations worldwide. For example, prior to the outbreaks in Egypt, the country was reported to export 180 million day-old-chicks plus 500,000 mature fowl a year. Almost 12 million live chickens were officially imported into the Ukraine in 2004 and more than 16 million into Romania. In Turkey, one factory has the capacity to produce over 100 million hatching eggs per year, many of them exported to Eastern Europe and the Middle East. Recent outbreaks in India, Nigeria and Egypt originated within the poultry industry, and there is strong circumstantial evidence that movements of poultry and poultry products are responsible.

For obvious reasons, little information is available on the extent of the unregulated and illegal poultry trade. However, recently it was revealed that poultry meat is being illegally imported from Asia into the United States; in October 2005, 3,000 chickens were intercepted by Italian customs officials after being smuggled into the country from China; and in November 2005 the UK authorities revealed that large quantities, possibly hundreds of tons, of chicken meat had been illegally imported from China, and fraudulently relabeled before being sold to food manufacturers across the country. In February 2006, 20 kilograms of chicken tongues from China were found by customs in Rio de Janeiro, Brazil, and 21 tons of (mainly) poultry meat from China were confiscated in southern Spain. These incidents indicate continuing lapses in border controls, despite the widely publicized risks. Illegal poultry movements are reported to be extensive in central Asia. In 2005, Ukraine's State Department of Veterinary Medicine said there had been substantial illegal re-exportation of meat from Ukraine to Russia from third countries.

Prevention and control

Better surveillance of wild birds, and study of the way that the virus behaves in wild bird populations, are very important. But it is even more important that preventive measures for H5N1 concentrate on better bio-security - surveilling and testing poultry, controlling the movements and sale of poultry, poultry products and cage birds, regulating the use of poultry manure used in aquaculture and agriculture, and stepping up national and international efforts to control the illegal trade in poultry, poultry products and captive wild birds.

The role of wild birds must be seen in the much larger context of the global poultry industry and the movements of huge quantities of poultry products around the world. Focusing on wild birds alone is misplaced and a potentially dangerous diversion of energy, effort and resources. Attempts to cull migratory wild birds or destroy their habitat are highly misguided - experience shows that this approach is completely ineffective, and indeed is likely to make matters worse.