高级检索

    黄土高原生态系统健康时空演变格局及水文循环驱动分析

    Spatiotemporal evolution of ecosystem health and its hydrological driving mechanisms in the Loess Plateau

    • 摘要:
      目的 探究黄土高原不同生态系统健康时空演变格局及其对水文循环要素的非线性响应机制,为区域生态恢复与水资源管理提供参考。
      方法 以黄土高原农田、森林和草地生态系统为对象,基于“活力—组织力—弹性—生态系统压力”(VORS)框架评估其健康状态,结合传统与空间马尔科夫模型揭示动态演变规律,并利用XGBoost-SHAP模型解析水文循环要素对生态系统健康的具体影响。
      结果 (1)生态系统健康整体提升。其中,森林生态系统高等级区增幅最大,占比增加35.52%,由中部向东部扩展;草地生态系统低等级区显著缩减,占比降低25.60%,向西部收缩。(2)生态系统健康的邻域效应显著,低等级区具负向溢出效应,抑制邻域健康等级向上转移;高等级区具有正向溢出效应,促进邻域健康等级提升。(3)农田和草地生态系统对水文要素变化敏感,降水为核心主导因子,对两类生态系统健康的贡献率分别为38.80%和45.80%。森林生态系统则更依赖水文要素的整体耦合作用,其交互效应总和约为单因子主效应的1.18 ~ 1.43倍。
      结论 本研究揭示了黄土高原不同生态系统健康水文驱动机制的显著差异,以及植被恢复背景下生态系统与水文复杂的非线性关系,对区域分区治理和可持续发展具有重要意义。

       

      Abstract:
      Objective Exploring the spatio-temporal evolution of ecosystem health and its nonlinear response to hydrological elements on the Loess Plateau is crucial for optimizing regional ecological restoration and water resource management.
      Method Focusing on farmland, forest, and grassland ecosystems on the Loess Plateau, this study assessed ecosystem health using the "Vigor-Organization-Resilience-Ecosystem Stress" (VORS) framework. We employed both traditional and spatial Markov models to reveal their dynamic evolution patterns and utilized the XGBoost-SHAP model to analyze the specific impacts of hydrological cycle elements on ecosystem health.
      Result (1) Ecosystem health improved significantly overall. Forest ecosystems showed the most pronounced increase in high-health areas (proportion increased by 35.52%), expanding eastward from central regions. Grassland ecosystems exhibited the most significant reduction in low-health areas (proportion decreased by 25.60%), contracting westward. (2) Low-health areas exerted a negative spillover effect, reducing the probability of neighboring ecosystems transitioning to higher health levels. Conversely, high-health areas demonstrated a positive spillover effect, promoting health level upgrades in adjacent areas. (3) Farmland and grassland ecosystems were highly sensitive to changes in hydrological elements, primarily dominated by precipitation (contributing 38.80% and 45.80% to their health changes, respectively). Forest ecosystem health, however, relied more on the synergistic coupling of multiple hydrological elements, where the combined interaction effects were approximately 1.18 to 1.43 times the main effects of individual factors.
      Conclusion This study reveals significant differences in the hydrological driving mechanisms for the health of distinct ecosystems on the Loess Plateau and elucidates the complex nonlinear relationship between ecosystems and hydrology within the context of vegetation restoration. It provides a theoretical basis for regional zoning management and sustainable development.

       

    /

    返回文章
    返回