Investigating the underlying drivers of vegetation dynamics in cold-arid mountainous

• Published in: Catena (Giessen) Vol. 237; p. 107831
• Main Authors: Xiao, Xiong, Guan, Qingyu, Zhang, Zepeng, Liu, Hanqi, Du, Qinqin, Yuan, Tingwei
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 30.03.2024
• Subjects: Boosted regression tree model; Cold-arid mountainous region; Driving mechanisms; Structural equation model; Vegetation changes; Vegetation changes; Driving mechanisms; Boosted regression tree model; Structural equation model; Cold-arid mountainous region
• ISSN: 0341-8162; 1872-6887
• DOI: 10.1016/j.catena.2024.107831

•A system for vegetation change mechanisms in cold-arid mountainous was developed.•Climate factors dominate vegetation greening and exist nonlinear thresholds.•Topographic factors coupled with multi-factors influence vegetation degradation.•Reasonable sustainable planning can avoid further degradation of vegetation. Understanding vegetation evolution trends and their driving mechanisms are essential to uncover changes in ecosystem structure and function. Under the interaction of natural and human factors in cold-arid mountainous region, the main controlling factors and degree of nonlinear influence are not yet clear. This study utilized trend analysis, along with modeling techniques such as boosted regression trees (BRT) and structural equation model (SEM), to identify and quantify the contributions and nonlinear response thresholds of natural and human factors to normalized difference vegetation index (NDVI) in the Qilian Mountains from 2000 to 2020. The results showed that the implementation of ecological projects made the overall vegetation into greening trend (91.06%), while irrational human activities led to degradation in the central-eastern regions. The BRT showed that climatic factors dominated for overall vegetation changes, greening and significant greening. Thresholds exist for different drivers to influence both greening and degradation of vegetation. Regarding degradation/significant vegetation degradation, the importance of climate factors decreased, and topography ranked first, followed by population density and LUCC. The expansion of building and unused land will inhibit vegetation greening. SEM revealed that climate factors continued to maintain positive effect with the warm-humid change. The destruction of forestlands and grasslands led to significant degradation of vegetation in the central-eastern regions with the largest total negative effect. Topographic factors coupled with multi-factors inhibited vegetation growth in the central-eastern regions. This study provided a new framework for accurately assessing the dynamic evolution of vegetation, and better reveal the multi-factor driving mechanism vegetation changes in cold-arid mountainous regions. Future research should consider combining more accurate AI algorithms with field surveys to quantify the coupling mechanisms between factors and vegetation dynamic. It is expected to provide scientific guidance for vegetation ecosystem planning in other mountainous regions globally and support the achievement of sustainable development.