研究生知识库

中国南方喀斯特地区（总面积约1.94×106 km2），在全球三大喀斯特集中分布区中连片裸露碳酸盐岩面积最大，是青藏高原隆起在东亚大陆亚热带气候区形成的一个生态脆弱性高、海拔梯度大、地势格局复杂的独特地理单元。然而，喀斯特地区生态系统脆弱，极易发生干旱和土地退化，加之区域内人口密度大，人类活动剧烈，导致生态环境恶劣，生态系统恢复工作进展困难。喀斯特生态系统特殊且脆弱，其生境与相同气候条件下常态地貌上地带性常绿阔叶林有明显的不同，其地质背景下发育的喀斯特土壤顶级森林植被以及其他植被类型具有独特的现存格局和动态演化特征。总体上看，中国南方喀斯特地区植被与环境关系研究大多在小尺度上展开，而在中尺度和大尺度上多以净初级生产力（NPP）和遥感观测的归一化植被指数（NDVI）作为植被表征而进行植被与环境关系分析，植被类型与环境因素的相互作用尚未被系统量化。因此，确定区域自然环境异质性，系统而准确地揭示中国南方喀斯特地区不同尺度上影响植被分布的主导环境因素，可以为我们探寻植被景观发展演化的过程和机制提供基础，为区域生物多样性保育和植被恢复工作，以及区域植物资源的开发利用和保护提供指导。为此，我们分析了中国南方喀斯特地区7种植被大类中的30种植被亚类的分布数据，以确定和量化与植被分布变化相关的环境、生物和人为干扰因素，并研究这些因素如何随尺度（地理范围）而变化。 本文分别从大尺度（整个区域）、中尺度（喀斯特和非喀斯特区域）和小尺度（小流域和群落水平）上分析了中国南方喀斯特地区植被与环境因子的关系。在尺度划分和直接梯度分析的概念设置内，我们对主要植被类型形成的地理范围与其物理环境的潜在格局之间的关系进行了全面的空间明确的分析。总体上，植被的分布可以用环境因素来解释，多种物理环境因子协同控制不同植被类型的空间分布。在不同尺度上，气候和地理变量分别占据主导地位。植被格局中平均约1/3的可解释变异可以通过环境因子的空间结构来解释。主要结果如下： 1、成功地利用梯度分析探寻出了不同空间尺度范围内控制植被分布的主要因子。在中国南方喀斯特地区大尺度上（同时包含喀斯特和非喀斯特），气候仍是植被分布的主控因子，地貌土壤因子作为协变量，对植被分布也具有重要影响；中尺度上，在非喀斯特地区，分析结果与大尺度结果类似，气候因子（占TVE的70%）仍是主控因子，而地貌土壤因子（占TVE的40%）作为次要因子协同影响植被分布。此外，非喀斯特地区TVE达到了0.31，超过大尺度上的总解释度。然而，在喀斯特地区，地貌土壤因子取代气候成为主控因子，气候因子作为协变量对植被分布产生影响。小尺度上，受岩石裸露率和坡位共同限制的土壤水分是植被分布的主要影响因素。 2、不同空间尺度的分析结果表明，环境因子对植被分布的解释度从21%-65%不等，这一结果与多数大尺度梯度分析表现一致；地貌土壤因子倾向于在中小尺度范围上或喀斯特地貌广泛分布的空间范围内表现出更强的解释性；气候因子在较大尺度或非喀斯特地貌广泛分布的空间范围内表现出更强的解释性；人为干扰因子在较小尺度下作为协变量参与对植被分布的解释，作用显著但并不突出；大部分土壤理化属性在多个尺度下均未表现出对植被分布的显著影响，仅土壤深度一个指标在多个尺度下均出现，且影响作用颇大。 3、梯度分析模型解释度随取样数量的降低而呈现下降趋势；但随着尺度的进一步缩小，在一定范围内，模型的解释度又逐渐上升；在小尺度空间范围内，模型的解释度受取样数量的影响不再明显，而取决于取样的精度是否能够满足对小区域内生态过程的描述。小尺度范围上，地貌土壤因子作用明显，同时群落内的种内种间竞争也成为控制植物分布的重要影响因素。 研究结果对区域保护战略中的国土空间规划、生物多样性保护、气候变化应对、区域生态监测清单和方案设计等具有重要的指导意义。

The South China Karst (with a total area of about 1.94×106 km2) is the largest area with exposed carbonate rock among the three major karst concentrated areas in the world. It is a unique geographical unit with high ecological vulnerability, high altitude gradient and complex topographic pattern formed by the uplift of the Qinghai-Tibet Plateau in the subtropical climate region of East Asia continent. However, the fragile ecosystem in karst area is prone to drought and land degradation. In addition, the high population density and intense human activities in the area lead to the poor ecological environment and difficult progress in ecosystem restoration. The karst ecosystem is special and fragile, and its habitat is obviously different from the zonal evergreen broad-leaved forest on the normal landform under the same climatic conditions. The karst forest vegetation and other vegetation types developed under its geological background have unique existing pattern and dynamic evolution characteristics. In general, most of the vegetation-relationship studies in the South China Karst are carried out on a small scale. On the mesoscale and large scales, net primary productivity (NPP) and normalized difference vegetation index (NDVI) are often used as vegetation characterization to analysis the vegetation-environment relationship. However, the interaction relationships between vegetation types and environmental factors have not yet been systematically quantified. Therefore, determining the heterogeneity of regional natural environment, systematically and accurately revealing the key environmental factors affecting vegetation distribution at different scales in the South China Karst, can provide a basis for us to explore the process and mechanism of vegetation changes, and provide a guidance for regional biodiversity conservation and vegetation restoration, as well as the development, utilization and protection of regional plant resources. To this end, we analyzed the distribution data of 30 vegetation types in 7 groups of vegetation type in the South China Karst to identify and quantify the environmental, biological and human disturbance factors associated with the changes in vegetation distribution, and to study how these factors vary with the scales (geographical range). In this paper, the relationship between vegetation and environmental factors in the South China Karst was analyzed from the large scale (whole region) to mesoscale (karst and non-karst regions) and small scale (small watershed and community level). Within the conceptual settings of scale partitioning and direct gradient analysis, we conducted a comprehensive and spatially explicit analysis of the relationship between the geographical extent of major vegetation types and the potential pattern of their physical environment. In general, the distribution of vegetation can be explained by environmental factors, and the spatial distribution of different vegetation types can be controlled by various physical environmental factors synergistically. At different scales, climatic and geographical variables dominate, respectively. An average of about 1/3 of the interpretable variation in the vegetation pattern can be explained by the spatial structure of the environmental factors. The main results are as follows: 1. The main factors controlling vegetation distribution in different spatial scales were successfully explored by gradient analysis. On a large scale (including both karst and non-karst regions) of the South China Karst, climate is still the key factor controlling the vegetation distribution. Geomorphic and soil factors, as covariables, also have a significant impact on vegetation distribution. On the mesoscale, in the non-karst areas, the analysis results are similar to those of the large-scale. Climatic factors (accounting for 70% of TVE) are still the key factors, while geomorphic and soil factors (accounting for 40% of TVE) act as the secondary factors to affect vegetation distribution synergistically. In addition, TVE in non-karst areas reached 0.31, which exceeded the total interpretation on the large scale. However, in karst areas, geomorphic and soil factors replaces climatic factors as the key factors, and climate factors, as a covariate, has a weaker impact on vegetation distribution. On the small scale, soil moisture, which is limited by both rock exposure rate and slope position, is the key factor affecting vegetation distribution. 2. The results of CCA analysis at different spatial scales showed that the interpretation of vegetation distribution by environmental factors varies from 21% to 65%, which is consistent with most large-scale gradient analyses. Geomorphic and soil factors tend to be more explanatory in the range of medium and small scales or in the spatial range where karst landforms are widely distributed. Climate factors are more explanatory in large scale or in the spatial range where non-karst landforms are widely distributed. As covariates, human disturbance factors participate in the interpretation of vegetation distribution on a small scale, but their effects are significant but not outstanding. Most of the physical and chemical soil properties have no significant effect on vegetation distribution at multiple scales, and only one index of soil depth appears in multiple scales with a significant influence. 3. The interpretability of the gradient analysis model showed a declining trend with the decrease of the number of samples, however, with the further reduction of the scale, the interpretation of the model increased gradually within a certain range. In the small scale space, the interpretation of the model is no longer significantly affected by the number of samples, but depends on whether the accuracy of sampling can meet the description of ecological processes in this small area. At the small scale, geomorphic and soil factors play a significant role on plant distribution, while interspecific and intraspecific competition within the community also becomes an important factor controlling plant distribution. The results of this study have important guiding significance for the land spatial planning, biodiversity conservation, climate change impact, regional ecological monitoring and other regional protection strategies.