%0 Journal Article
%A Chi-Cheng Liao
%A Huan-Yu Lin
%A Su-Wei Fan
%A Yi-Huey Chen
%T Virtual species as baseline for predicting trends of real species distributions in mountainous regions
%D 2025
%J Taiwania
%V 70
%N 4
%P 658-676
%U https://taiwania.ntu.edu.tw/abstract/2124
%X Species distribution models (SDMs) play a crucial role in predicting species' geographic distributions, especially in mountainous regions where complex terrains pose significant challenges for sampling efforts. To enhance the accuracy of SDM predictions, researchers have employed high-resolution climate datasets derived from meteorological data. These datasets were used to simulate five virtual species across different climatic niches using principal components analysis (PCA). This approach allowed for a thorough examination of various environmental conditions, such as coastal, inland, and leeward slopes, thereby capturing the intricate climatic heterogeneity within mountainous landscapes. In these simulations, eleven algorithms were assessed using the Expected Fraction of Shared Presences (ESP), which measures the overlap between simulated suitable ranges and predicted potential ranges. Ensemble modeling proved superior, particularly when dealing with larger sample sizes, showing minimal impact from geographical distribution variations. However, smaller sample sizes notably reduced ESP values, geographical distance among sample points provide signals of niches that will be identified through accurate predictions of ensemble modeling. Temperature has been determined to be more critical than precipitation for species distribution, heavily influenced by elevation gradients. Climate change projections suggest that by the 2050s and 2090s, species currently found in inland areas or on leeward slopes may experience range expansions. Conversely, species inhabiting coastal areas or windward slopes are projected to face range restrictions or local extinctions. This study supports the development of targeted conservation strategies in topographically complex areas, offering scientific support for conservation planning in mountainous regions and forecasts of species distributions under future climate change.
%M doi:10.6165/tai.2025.70.658