Mapping small mammal optimal habitats using satellite-derived proxy variables and species distribution models

Area of Science:

• Ecology
• Remote Sensing
• Biodiversity Studies

Background:

• Small mammals significantly influence ecosystems, affecting vegetation, soil, and food webs.
• Population cycles of small mammals can lead to agricultural impacts and disease transmission.

Purpose of the Study:

• To identify key landscape variables influencing small mammal distribution and abundance.
• To develop species-specific distribution models for better habitat characterization.

Main Methods:

• Utilized Earth observation data and in-situ surveys for nine small mammal species.
• Generated random forest species distribution models (SDMs) for Narati, China, and Sary Mogul, Kyrgyzstan.
• Quantified landscape characteristics and dynamics impacting species life cycles.

Main Results:

• Identified species-specific landscape drivers, with variable importance ranging from 3 to 26 metrics.
• Predicted higher abundances in grasslands (e.g., Microtus obscurus), forests (e.g., Myodes centralis), and mixed/riparian areas (e.g., Apodemus uralensis).
• Random Forest models achieved high validation (R² 0.670–0.939), identifying abundance hotspots.

Conclusions:

• Landscape dynamics and habitat types are crucial for small mammal distribution and abundance.
• Species-specific SDMs provide a more accurate, species-defined perspective on optimal habitat.
• Earth observation data effectively enhances understanding of ecological linkages to small mammal populations.