Trekking through the Quinling Mountains in southwestern China is no holiday: the terrain is remote and rugged, and people have died falling from the slopes. The natives aren't particularly welcoming either. The region's giant pandas aren't the friendly sorts they are sometimes portrayed to be, being carnivores, and although there are no reports of them eating humans, there is anecdotal evidence that they have taken livestock.

Nevertheless, groups of researchers and students from Chengdu Research Base of Giant Panda Breeding, make frequent trips into this harsh terrain. Their aim: to collect giant panda droppings. This sounds like a strange hobby - particularly in view of the risks - but if it isn't some dangerous new sport for thrill seekers, what are they up to?

Professor Stephen Kemp and Dr Phill Watts at the University of Liverpool explain. "It's part of a global collaboration to prevent the giant panda from becoming extinct," says Dr Watts. "Chengdu Research Base cares for over 30 pandas raised through a captive breeding programme. But breeding isn't just about increasing the numbers of giant pandas per se. It is likely that inbreeding will have severe consequences, and maintaining genetic diversity is crucial to the success of the giant panda breeding and release programme."

Maintaining genetic diversity is problematic: in captivity, fertilisation of female pandas is ensured by artificial insemination from several males, in addition to natural copulation, so the true paternity of cubs is never known. Moreover, very little is known about the genetic structure of pandas in the wild.

To address this problem, Dr Shen Fujun, a zoologist at Chengdu, received a Wellcome Trust travel grant to visit the University of Liverpool. In collaboration with Professor Kemp and Dr Watts, he received training in DNA marker development techniques not yet widely established in China, allowing him to establish a panel of genetic markers specifically for the giant panda.

"We needed blood samples initially because you need high-quality DNA to establish libraries of markers," explains Professor Kemp. "What we aimed to do was to make robust, reliable markers that give clear answers for degraded DNA. So once Dr Shen went back to China, he didn't need to collect any more blood: using the markers he could type panda DNA from faeces or hair. It's not only more comfortable for the pandas, it's probably the only way they can realistically be studied in the wild."

With Dr Shen, the team at Liverpool also established protocols for the extraction of DNA from faeces. "You have to dry out the faeces, either by using specific buffers, freezing or treating them with ethanol to preserve the DNA," says Dr Watts.

These protocols, and the libraries of genetic markers, will make it possible for researchers to track the movements and breeding success of giant pandas both in the wild and captivity - without any discomfort to the pandas - to establish the number of individuals in the wild, the paternity of cubs, trace family trees, and develop appropriate breeding strategies to maximise genetic diversity.

Dr Shen Fujun will also use the experience gained in Liverpool to facilitate the training of other Chinese scientists in the technique - and to develop markers in other endangered species in China, including the crested ibis, the red panda, several Asian cat species - even the South China Tiger. Despite the fact it was declared extinct in 2002 by a team of US scientists, Chinese researchers still fiercely believe the cat is still at large and have offered a large reward for recent photographs proving its continued existence.

See also

• Animal Genomics Laboratory at the University of Liverpool: Further details of the giant panda research project with Dr Shen Fujun
• Chengdu Research Base of Giant Panda Breeding
• Chester Zoo and the North of England Zoological Society support the research