Machine Learning Model Analysis of Breeding Habitats for the Black-necked Crane in Central Asian Uplands under Anthropogenic Pressures

Abstract

Abstract The black-necked crane (Grus nigricollis) is the only alpine crane species and is endemic to the Tibetan Plateau. The breeding habitats of this species are poorly understood, which greatly hampers practical research and conservation work. Using machine learning methods and the best-available data from our 7,000-kilometer mega-transect survey and open access data, we built the first species distribution model (SDM) to analyze the black-necked crane’s breeding habitats. Our model showed that current conservation gaps account for 26.7% of its predicted breeding habitats. Specifically, the northern parts of the Hengduan Mountains and the southeastern Tibet Valley, the northern side of the middle Kunlun Mountains, parts of the Pamir Plateau, the northern Pakistan Highlands and the western Hindu Kush should be considered as its main potential breeding areas. Additionally, our model suggested that the crane prefers to breed in alpine meadows at an elevation over 2,800 m, a maximum temperature of the warmest month below 20.5 °C, and a temperature seasonality above 7,800 units. The identified conservation gaps and potential breeding areas can aid in clearly prioritizing future conservation and research, but more attention and study should be directed to the unassessed Western Development of China to secure this endangered crane lineage and other wildlife on the Tibetan Plateau.