Herders key figures in plateau wildlife conservation

With support from the national park administration and local governments, team members train herders and ecological rangers to deploy infrared cameras and monitor wildlife.

The work is built around local knowledge. Herders know where snow leopards appear, which slopes animals use and where tracks or droppings are likely to be found.

"They are local scientists," Li said. "We bring modern methods and work with them to produce monitoring results."

The team divided the area into 72 grids and works with 93 herders and rangers to maintain 98 infrared cameras. The network has accumulated about 9 terabytes of data and over 1,900 snow leopard images.

Since three snow leopards were first recorded in 2015, the team has identified 142 individuals, with about 20 to 30 regularly detected in the valleys on an annual basis.

It has also recorded 17 leopards, including a female believed to be at least 12 years old that has given birth three times and raised five cubs.

Blue sheep density in Angsai is about four per square kilometer, providing sufficient prey for large carnivores and indicating a relatively intact food chain, Li said.

Technology has accelerated the work. An online database allows volunteers to identify animals remotely, while artificial intelligence filters empty frames, cutting manual screening by about 40 to 50 percent. But rain, snow and thick fur can blur a snow leopard's markings.

"AI can help sort similar images, but final identification still depends on experienced volunteers and researchers," Li said.

However, technology cannot replace the people who know the land.

Herders, Li said, taught her to recognize snow leopard scrapes and animal droppings, and guided researchers collecting fecal samples for genetic studies.

"Without local herders and rangers, we could not have maintained 11 years of continuous monitoring," she said.