Researchers suggest the disease is caused by African elephants transmitting the infection to Asian elephants in captivity.

The first known victim of the virus was a 16-month-old Asian elephant, the first ever born at the National Zoo in Washington DC. Its death in 1995 baffled zookeepers.

Since then seven other Asian elephant deaths have been associated with the virus. In another two cases, where records are incomplete, there is also a high probability of the same infection.

The infection kills baby Asian elephants soon after they are born, causing internal bleeding and heart failure. It hits suddenly, killing in just a few days.

The cells that line the blood vessels, the heart and other organs are attacked, resulting in the fatal haemmorrhaging.

The viral strain was identified by researchers from the Johns Hopkins School of Medicine, Baltimore, and Washington's National Zoological Park. They have published their findings in Science Magazine.

Non-lethal herpes

They used a type of DNA analysis - consensus primer polymerase chain reaction (PCR) - combined with sequencing to make their discovery.

It was found that the same virus was present in many African elephants in Zimbabwe and South Africa, both in zoos and in the wild.

But apart from the usual blisters associated with herpes, the virus was clearly non-lethal.

The research scientists voiced their concerns for the elephant populations in zoos.

African elephant - may carry virus, but no lethal effects

"It is likely that the virus was transmitted from the African to the Asian elephants in the zoos," said Dr Laura K Richman.

"That's the only way we can account for the same virus being present in both populations."

Fellow scientist Gary Hayward added: "This is very troubling because ... there may still be carrier African elephants in zoos."

He said that blood tests were needed to identify which elephants may be carrying the virus. Separating Asian and African elephants could prevent more deaths, he added.

Hope of cure

But the future for the elephants is not completely doomed. In two cases a treatment of famciclovir, an antiviral drug, proved successful.

"We were able to cure these elephants, which is promising. If caught early, the infection appears treatable," Dr Richman said.

A similar disease has also been found in African elephants, but a different, though related herpes virus was associated with the infection.