Modeling distribution and habitat suitability for snow leopard (Panthera uncia) in the Great Gobi-A Strictly Protected Area of Mongolia

Ing. B.Sc. Juno Shimada, BSc

Diplomová práce

Abstract:

Understanding species distribution and habitat suitability is critical for conservation efforts, especially for elusive species like the snow leopard (Panthera uncia). Species Distribution Models (SDMs) are effective tools for predicting habitat suitability using species occurrence records and environmental variables, even for elusive animals. This study aimed to provide a prediction of snow leopard habitat suitability in Mongolia and northern China and to compare the effectiveness of different occurrence record collection techniques: camera traps, GPS collars, and their combination. Using Species Distribution Models (SDMs), we analyzed 252 occurrence records from GPS collars and 17 from camera traps, alongside 10 environmental variables including 4 bioclimatic variables, 4 topographic variables, and 2 prey species distribution (Siberian ibex Capra sibirica and Argali sheep Ovis ammon), through an ensemble model framework which combines the outputs from multiple SDMs. The ensemble models identified 25260 km2 (Camera traps), 31513 km2 (GPS collar), and 30220 km2 (Combined) as very high and high suitable area, which is less than 0.6% of the total study extent across all techniques. Suitable habitats were mainly located in the GGASPA and the Altai Mountain range in western Mongolia, with additional suitable areas near northern Qinghai province, China. These suitable habitats of all techniques were characterized by a prey distribution, particularly Argali sheep. The GPS collar and Combined models outperformed the Camera traps model, with the Camera traps model showing fewer unsuitable areas, while the GPS collar and Combined models were more conservative, indicating highly suitable habitats were confined to specific regions. This study provided a first prediction of suitable habitat for snow leopard in Mongolia and northern China, but also highlighted the need for various improvements. These findings emphasize the need for more comprehensive data collection in broader regions. The results also suggested that to use a combination of both data types: GPS collar and camera traps, while being mindful of potential geographic biases.

Keywords

ecological niche modeling, remote sensing, prey abundance, camera trap, GPS collar, China

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 16. 9. 2024

Obhajoba závěrečné práce

• Obhajoba proběhla 12. 9. 2024
• Vedoucí: doc. Francisco Ceacero Herrador, Ph.D.
• Oponent: Vítězslav Moudrý, Ph.D.

Česká zemědělská univerzita v Praze

Fakulta tropického zemědělství

Magisterský studijní program / specializace:
Wildlife and Livestock Production, Management and Conservation / Wildlife Management and Conservation