Pig surveillance is largely the remit of animal-health organizations, agriculture ministries and the farming industry. Their main concern tends to be that any reports of the pandemic virus in pigs might provoke overreactions such as the mass culling of pigs that took place in Egypt, or trade bans on pigs and pork. Within minutes of the World Health Organization (WHO) announcement on 11 June that swine flu had become a pandemic, Bernard Vallat, director-general of an intergovernmental trade body, the World Organisation for Animal Health (OIE), had reiterated that trade sanctions were unjustified. "So far the role of animals has not been demonstrated in the virus's epidemiology or spread," he asserted.

But some experts say that is an artefact of patchy to non-existent flu surveillance in pigs. In a paper published last week (G. J. D. Smith et al. Nature advance online publication doi: 10.1038/nature08182; 2009), Gavin Smith, a flu geneticist at the University of Hong Kong, and his colleagues concluded that "the lack of systematic swine surveillance allowed for the undetected persistence and evolution of this potentially pandemic strain for many years". The virus originated from a mixture of swine flu strains, and pigs are an "obvious" part of the epidemiology of the new virus, says Smith. Yet the number of swine-flu sequences in the international GenBank database is about a tenth of that for avian flu viruses. Circulation of the virus between pigs and humans is "definitely a possibility", he adds.

The pandemic virus has so far been found in pigs from just one farm, in Alberta, Canada, where it spread throughout the herd. But noone has been able to pin down how the herd became infected. Scientists at the Veterinary Laboratories Agency in Weybridge, UK, have shown that pigs can easily become infected with the virus, and readily transmit it between themselves and shed it into the environment. Past pandemic viruses have also gone on to become endemic in pig populations.

"It's absolutely surprising that a virus this contagious in both humans and swine, and which has been reported in humans in 76 countries, has only been reported in one swine farm in Canada," says Jimmy Smith, head of livestock affairs at the World Bank in Washington DC, and a member of the organization's flu task force. "It is highly likely that more pigs are infected in more places."

Absence of evidence of the pandemic virus in pig populations is not evidence of absence, concedes Steve Edwards, chairman of the OIE–FAO Network of Expertise on Animal Influenza (OFFLU), which coordinates work done by animal-flu surveillance labs worldwide, and former chief executive of the Veterinary Laboratories Agency. "But we should not assume it is there until we have evidence to say so," he insists.

Whereas flu surveillance has improved over the past six years in poultry and wild birds, pigs have been below the radar, says Ilaria Capua, an animal-flu expert at the Experimental Animal Health Care Institute of Venice in Legnaro, Italy. The avian H5N1 flu virus leads to serious disease in poultry and causes huge economic losses, so the OIE requires its 174 member states to report any outbreaks. In pigs, flu viruses, although common, tend to cause only mild disease, so there is no obligation to report cases of swine flu, much less take samples for genetic and antigenic analysis. The OIE has, however, asked its member states to voluntarily report any occurrences of the 2009 pandemic virus in pigs.

"Surveillance for swine flu is not something that has been high on the agenda of government services," says Mr Edwards. "It is seen as a farming-industry problem." Most flu surveillance in pigs is passive, relying on farmers or vets sending material to government labs. Active targeted surveillance with diagnostic tests is rarer, as it is more expensive.

OFFLU has called on labs worldwide to share what information they have on swine flu, and to sequence any samples they have obtained recently. So far, however, the response has been "limited", says Mr Edwards.

A meeting between experts from OFFLU and the WHO on 21 May — the conclusions of which were made public last week — recommended scaling up flu surveillance efforts in pigs, and OFFLU is now developing sampling and diagnostic protocols. Recommendations are all very well, but "somebody has to agree to fund all of this work", warns Mr Edwards.

Ironically, European Union funding for one of the world's largest pig surveillance networks expired in March, reports Naturenews. The European Surveillance Network for Influenza in Pigs, which was created in 2001, comprises nine European labs and one in Hong Kong. Although the network has not yet detected the new virus in pigs, its coordinator Kristien Van Reeth, an animal virologist at Ghent University in Belgium, admits that participating labs have taken just a few hundred to a thousand samples each over the past year. Network members hope that with the pandemic highlighting the need for better pig surveillance, new funding will be forthcoming.

It is now clear that the animal-and public-health communities underestimated the potential for pigs to generate a pandemic virus. Although pigs can be infected with many subtypes of flu, the three most common endemic strains are H1N1, H1N2 and H3N2. Most expected that any new pandemic would involve the introduction of a viral subtype not previously seen in humans, such as the avian H5 subtype, explains Capua. "The consensus was that a pandemic could not be caused by H1, H2 or H3 because the current human population would have antibodies against them," she says.

The emergence of the reassorted H1N1 pandemic virus — which current research indicates noone has any immunity to, apart, perhaps, from some people older than 60 — has changed that thinking. Moreover, Nature has learned that the international community was warned of such a risk in a presentation at a closed meeting between the OIE, the WHO and the Food and Agriculture Organization of the United Nations in Paris in February. In the presentation, the results of which are in press at the journal PLoS Pathogens, Capua showed that serum samples from people vaccinated against seasonal flu strains showed little or no cross reactivity against H1, H2 and H3 bird viruses, meaning that they would have no immunity.

This shows that the world needs a comprehensive surveillance system of all influenza subtypes and their evolution across many animal species, says Capua: "We should be looking at the bigger picture."