江河源区高寒草甸退化序列上"秃斑" 连通效应的元胞自动机模拟

高寒草甸草毡层"秃斑"是高寒草甸退化进程中最活跃的表征,明确其在退化序列上的演变时空规律是揭示草地退化动力学机制的关键之一。采用实地调研结合元胞自动机模拟的方法,对高寒草甸退化序列上秃斑格局动态进行了模拟,是对退化序列上秃斑的连续动态图谱表达,建立了高寒草甸由初步退化到系统崩溃的符合实际的草地秃斑连通的图谱序列,模拟吻合度达93.9%。通过对图谱序列关系的分析表明,秃斑连通进程分为3个阶段:低速连通期——第0—2年、跃变期——第2—7年、连通不可逆转期——第7—9年;最大秃斑面积、最大秃斑面积-秃斑总面积比的跃变过程,是草地退化等级的质变过程;通过对跃变期和连通不可逆转期相应的草地生态与恢复性能的对比分析,确定了连通阈值为54.5%;秃斑连通的过程伴随着临界阈现象的发生。由于连通不可逆转期"黑土滩"形成过程的不可逆性及形成后的巨大危害性,连通阈值的确定将为高寒草甸生态系统安全预警及其退化恢复治理提供依据。

Cellular automata simulation of barren patch connectivity effect in degradation sequence on alpine meadow in the source region of the Yangtze and Yellow rivers, Qinghai-Tibetan Plateau, China

The source region of Yangtze River and Yellow River plays an important role in the carbon source/sink cycle and is a focal point for ecological environment management; however, it is also one of China's most vulnerable ecological systems. This vulnerability is due to human and livestock population expansion; large area of alpine meadow are experiencing overgrazing and climate change, with barren patches developing and connecting with each other. Barren patches develop in a range of sizes and shapes, presenting an infertile "black soil" type landscape in the final stage of grassland degradation. The driving mechanisms of alpine meadow degradation are complex and remain controversial, making it difficult to determine individual causes of degradation and to take effective measures to counter them. Thus, this investigation employed "black box" theory to avoid these uncertain elements in its examination of spatial patterns and temporal evolution in barren patches. This investigation aimed to clarify the role barren patch evolution and connectivity plays in alpine meadow degradation, providing an increased understanding of alpine meadow ecosystems as the basis for ecological maintenance, restoration and management. A simulation method of complexity science, cellular automata, was used to model the development of barren patches in this study. The initial iteration data for the model were obtained from observations of moderately degraded meadows with the lowest barren patch percentage, and were used to draw a matrix in which each cell was represented as either 0 or 1; 0 representing a cell where vegetation cover was lower than 50% and 1 representing cells where vegetation cover was over 50%. Based on field observations of barren patch percentage and landscape structure, including the experience of local herdsman whose livelihoods are inextricably linked to the grassland, the simulation time step was set to one year per iteration. Further field observations of barren patch evolution and other features of their spatial distribution at different stages of degradation were used to define the rules of the model, and Matlab 7.0 was used as a platform for simulation. Cell neighborhoods were defined as the Moore type and cellular space was treated according to the reflective boundary rule. Using this combination of field research techniques and cellular automata simulation, realistic developmental graphs modeling the connectivity of barren patches from the initial degradation stage to the collapse of the alpine meadow system were established. The goodness of fit for the simulation analyses averaged 93.9%. Results further indicated that in the degradation sequence, there were three defined degradation stages: a low-speed connectivity stage from 0 to 2 years, a jump stage from 2 to 7 years and an irreversible connectivity stage from 7 to 9 years. Additionally, a sudden change was found to occur at the beginning of each stage, identifying a threshold characteristic in the process of barren patch connectivity. Through comparative analysis of the performances of grassland ecology and restoration between the jump stage and the irreversible connectivity stage, the connectivity threshold was calculated at 54.5% as a protection index of the alpine meadow. The irreversibility of ecological harm associated with large barren patches highlights the importance of determining and using the connectivity threshold to identify and determine priority sites for restoration.