When people think of bird flu in Asia, Cambodia rarely springs to mind. It is usually either the massive culling of birds in China or new outbreaks in Indonesia or Vietnam that grab the headlines.
Yet the H5N1 bird flu virus is widespread in Cambodia's poultry and wild bird populations alike. And with the country's limited health care capacity, inadequate surveillance and poor public health awareness, any investigation into bird flu here is a journey into the unknown.
In the face of these problems, the Cambodian government and various international organisations are keen to do everything they can to tackle the bird flu threat. The problem is where to start.
Ever since bird flu emerged in South-East Asia in 2003, wild birds have been seen as a possible reservoir for the virus. The country has some 540 species of wild bird, of which 160 are migratory.
The US-based Wildlife Conservation Society (WCS) is working with the Cambodian government and the Cambodian branch of the Pasteur Institute to assess the extent of H5N1 in wild bird flocks, which is not yet known.
"There has been limited surveillance, particularly in Cambodia in the poultry sector and even less in wild birds," says Martin Gilbert, field veterinarian for the WCS. "So far the only wild bird testing has been haphazard."
Detecting the virus in wild birds involves collecting faecal and blood samples, and taking throat swabs.
Gilbert explains that two different types of sample are taken. One keeps the virus alive, while the other allows the virus's genetic material to be stored.
The samples are taken to a new virology unit in the National Animal Health and Production Investigation Centre (NAHPIC), and are used to inoculate fertilised chicken eggs. The death of the developing embryo indicates further investigation is needed.
The suspected samples are immediately sent to the Phnom Penh Pasteur Institute, which can confirm whether a sample is positive within 24 hours.
Positive in the provinces
In October, NAHPIC analysed blood samples from 19 flocks of free-range domestic ducks in the south-eastern province of Prey Veng. Nearly one-quarter of the 267 samples were positive, indicating that the bird flu virus had been present in the area for several months. Similar results were also found for duck flocks in the adjacent province of Kampong Cham in July.
Yet despite this evidence that the virus has been circulating in Cambodia, the number of bird flu outbreaks reported is still low compared with neighbouring Vietnam.
Studies in Thailand and Hong Kong show that the wild bird species that are vulnerable to bird flu are mainly scavenging species, such as magpies and crows, which live in close proximity to people.
But in Cambodia, the potential for virus crossover between wild bird populations and domestic poultry is believed to be relatively small, partly because hunting has forced wild birds into areas where few people, or poultry, live.
"Cambodia can boast some of the last significant water bird colonies left in south-East Asia, but they are remote, and that's why they still exist," Gilbert says.
"The birds are nesting on emergent trees rising from the Tonle Sap lake. It is a flooded forest, which makes the nests particularly difficult to get to."
However, Cambodia's wild bird trade has created an 'artificial migration' by moving large numbers of non-migratory birds into areas where people live.
Birds for sale along the riverside
Much of this trade is driven by the Buddhist practice of merit giving, a time-honoured tradition of purchasing caged birds to set free – a way to earn celestial merit and ward off danger. Along the riverside in Phnom Penh, market stalls can be found full of caged wild birds, which people buy for less than US$1 to release as an offering at the local temples.
It is a regular practice at any time of year, but it is also linked to religious festivals. On major holidays there can be thousands of wild birds of many species concentrated in a small area.
Up to 16,000 birds can be sold in the space of two weeks, among them warblers, sparrows, swallows and munias, and several endangered species such as the Asian golden weaver — classified by the World Conservation Union (IUCN) as "globally near-threatened". Several of the species sold in cages are known to act as carriers of bird flu.
In Thailand, H5N1 has been isolated from the scaly-breasted munia, the wild bird species most commonly sold in cages. Thai authorities also detected H5N1 in tree sparrows, the third most common species at the markets.
"What has probably happened is that the birds have picked up the virus around human inhabitation," Gilbert says.
"A virus in a closed population quickly runs out of susceptible hosts and so burns itself out. However when you continuously feed new unexposed birds into the system you are effectively adding more fuel to the virus."
"The wild bird trade may actively promote the emergence of more virulent viruses," he says.
Any birds left behind at the end of the day are cooked and eaten. Children have been seen buying birds to eat raw. Although the sale of wild birds is illegal in Cambodia, the practice has strong traditional links with the Cambodian royal family. Gilbert explained that any change would have to come from a shift in the beliefs of the Khmer community itself.
The most effective way to identify an outbreak of bird flu is to monitor bird populations, but there are fears that the current surveillance systems are insufficient. They do not cover the whole of the country, only areas that have been identified as previous centres of outbreaks.
If an outbreak occurred in a province that is not being monitored, the whole system could fail — a problem that is being addressed by the United Nations Food and Agriculture Organization (FAO) and the Cambodian government.
The country's poultry industry is a central focus for surveillance, but it has had fewer cases of bird flu than neighbouring countries because it is much smaller.
Cambodia only has around 150 semi-commercial farms. Many have as few as 2,000-5,000 birds, while the largest contains 100,000 and is subject to strict bio-security.
Backyard chickens account for 90 per cent of the poultry stocks in the country, and owing to the way Cambodian villages are structured, domestic poultry populations do not intermix. Many domestic flocks are small and contained within private holdings, so it is not easy for a virus to spread between populations.
Routine checks on rural farms are problematic. The FAO has begun a pilot scheme for training local animal health workers to collect samples. But very often farmers refuse to allow them to take swabs because it stresses the laying ducks, making them lay fewer eggs, and putting the farmer at risk of losing income.
"When you are a farmer you face a difficult situation. If the test results are positive, they will want to kill all your ducks, and even if the result is negative you will lose a lot of money because for one week your ducks will stop laying," says Yves Froelich, a technical adviser for the FAO.
"When they suspect mortalities we ask them to impose immediate controls, but it is not so easy to convince the farmer," he says.
At present the Cambodian government does not compensate farmers who have had their poultry stocks either inspected or culled, and donors are hesitant to provide funding.
But the success of the FAO's animal health workers' pilot scheme is changing attitudes, and the programme is expanding into other provinces. By early 2006 the scheme will begin in 13 provinces deemed to be high risk areas.
The awareness gap
Unlike previous virus scares in Asia, bird flu is a predominantly rural problem. If H5N1 mutates, as many experts predict it will, an outbreak could occur far from healthcare centres.
It is no coincidence that all of Cambodia's human cases of bird flu to date were in Kampot, a province near the Vietnamese border: it is the Vietnamese health care system that is finding them.
In January this year, Cambodia's first human bird flu fatality, Tit Sokhan, died in Vietnam after seeking medical treatment there.
She sought medical help in Cambodia at first, but the doctor was unable to identify her illness. Later, health workers advised her to perform a traditional ceremony to appease the spirits.
This points to one of the main problems with identifying a bird flu outbreak — rural people are not aware of the virus.
Many villages are used to seeing birds suffer from seasonally high mortality rates. Often, entire populations of chickens are simultaneously wiped out in a village every year.
"The farmers we encountered earlier this year say they accept that chickens die this time of year, and they have always been safe to eat. They ask why there is a problem now," says Megge Miller, an epidemiologist for the World Health Organization.
"We are not asking people to do something simple. We are asking them, in the face of their poverty, not to eat chickens that have died," she says.
"In that respect, it is going to take a long time to really change that behaviour, particularly if chicken is the sole source of protein for the family, or an important source of income," she told SciDev.Net, "It is money they cannot afford to waste."
Slaughtered chickens on sale in a Kandal market
But education is only part of the solution. Cambodia lacks drugs and health care, and has a poor infrastructure, severely limiting its ability to contain an outbreak. The country desperately needs more funds to combat H5N1, and donors are starting to realise the extent of the problem.
The United States recently agreed to give Cambodia nearly US$2 million to fight a potential human flu pandemic. The measures include training rapid response teams, developing Cambodia's capacity to safely and accurately identify bird flu, and creating a national stockpile of medicines.
Ly Sovann, deputy director of Cambodia's Centre for Disease Control, reported that the country's Ministry of Health has just over 100 treatment courses of Tamiflu (oseltamivir) available, as little as one course for each province.
He believes that at least 20 courses are needed in each province to be able to adequately deal with an outbreak, and that even then it will depend on the speed at which an outbreak can be identified.
"We do not know whether we will pick it up early enough to have only a family cluster, or if the system will pick it up too late when the virus has spread to village level," says Sovann.
"We have shown with malaria, SARS, and TB that we have experience in combating dangerous diseases. What we need is the support."
Back at the riverfront, work is at hand to prepare for the annual water festival. The celebration marks the end of the rains and will see up to a million people travelling to the capital from the provinces. No one knows how many poultry or wild birds will be brought with them.
"If we can separate different species of poultry from each other, stop them living beside humans, stop humans moving them and stop them mixing with wild birds on rice paddies, then we can solve the problem," says Martin Gilbert of the WCS. "We need to break the chain of transmission."
As always, and in Cambodia particularly, this is easier said than done.