Giant panda distribution ranges in the Liangshan Mountains

Comprehending the population trend and understanding the distribution
range dynamics of species is necessary for global species protection.
Recognizing what causes dynamic distribution change is crucial for
identifying species’ environmental preferences and formulating protection
policies. Here, we studied the rear-edge population of the flagship
species, giant pandas (Ailuropoda melanoleuca), to 1) assess their
population trend using their distribution patterns, 2) evaluate their
distribution dynamics change from the 2nd (1988) to the 3rd (2001) surveys
(2–3 Interval) and 3rd to the 4th (2013) survey (3–4 Interval) using a
machine learning algorithm (The Extremely Gradient Boosting), and 3)
decode model results to identify driver factors in the first known use of
SHapley Additive exPlanations. Our results showed that the population
trends in Liangshan Mountains were worst in the 2nd survey (k = 1.050),
improved by the 3rd survey (k = 0.97), but got worse by the 4th survey (k
= 0.996), which indicates a worrying population future. We found that
precipitation had the most significant influence on distribution dynamics
among several potential environmental factors, showing a negative
correlation between precipitation and giant panda expansion. We recommend
that more study is required to understand the micro-environment and animal
distribution dynamics. We provide a fresh perspective on the dynamics of
Giant Panda distribution, highlighting novel focal points for ecological
research on this species. Our study offers theoretical underpinnings that
could inform the formulation of more effective conservation policies.
Also, we emphasize the uniqueness and importance of the Liangshan
Mountains giant pandas as the rear-edge population, which is at a high
risk of population extinction.