Abstract:

Ancient trees, as unique elements of urban and rural landscapes and witnesses to historical changes, carry rich cultural connotations and represent non-renewable green cultural heritage. They have important ecological value and serve as key resources for promoting rural revitalization and shaping the village landscape. The unique landforms and soil characteristics of the Qiongbei volcanic lava area have nurtured a rich and distinctive ancient tree community, adding unique charm to the ecological environment and cultural heritage of the region. This study aimed to analyze the spatial distribution pattern and influencing factors of ancient trees in the Qiongbei volcanic lava area. The results show that the spatial distribution of ancient trees is influenced by both natural and human factors, among which soil texture is the most dominant. In particular, in the volcanic lava landform area, the difference in soil type has a significant impact on the growth of ancient trees. The diverse microtopography of terraces, slopes, and depressions in the volcanic hilly area provides a suitable environment for the growth of ancient trees owing to differences in slope, light, humidity, and wind speed. Shady, humid, and sheltered areas are particularly suitable for the growth and preservation of ancient tree species. Simultaneously, microclimatic changes caused by terrain undulations are an important factor affecting the distribution of ancient tree species. In terms of human factors, population density and land use intensity showed a certain “promoting effect” on the distribution of ancient tree species. The main objectives of this study are to quantitatively analyze the spatial distribution pattern of ancient trees based on the ancient tree resource survey data in the area using Geographic Information Systems (GIS) and the Geographically Weighted Regression (GWR) model and to comprehensively discuss its aggregation characteristics and influencing factors. The study analyzed the global spatial distribution of ancient tree points through nearest neighbor index (NNI) analysis to determine the distribution pattern of ancient trees and found that it showed significant clustered distribution characteristics. Furthermore, using kernel density analysis (KDE) and hotspot analysis, the cold and hot spots of ancient trees were identified, and the changes in the density of ancient tree points were quantified. Through grid multi-scale debugging, a spatial scale of 1 km×1 km was finally determined for a comprehensive analysis of the influencing factors. The study selected 11 factors, including altitude, rainfall, temperature, sunshine, soil type, soil texture, soil organic matter content, population density, per capita GDP, land use intensity index, and distance from traditional villages, as influencing factors for analyzing the spatial distribution of ancient trees. To effectively eliminate the interference of multicollinearity on the analysis results, this study used the Ordinary Least Squares (OLS) regression model in the ArcGIS spatial statistical analysis tool. This model, based on the principle of least squares, estimates the model parameters by minimizing the residual sum of squares between the observed and model-predicted values, thereby screening out the most representative independent variables. Finally, nine independent variables were selected for the comprehensive analyses. By calculating the mean of the relevant variables in each grid and performing a spatial regression analysis, this study used the GWR model to reveal spatial heterogeneity and quantitatively explain the impact of each variable on the changes in the dependent variable at different spatial locations. The research results can be summarized as follows: 1) Richness and distribution characteristics of ancient tree resources: The study found that the area is home to a rich diversity of ancient tree species, with 521 trees from 15 species and eight families. Of these, 95.59% are between 100 and 300 years old, indicating that the majority of ancient trees in this area are middle-aged or older, which enhances the ecological and cultural values of the area. The tree species were mainly Moraceae, with the most common being Ficus microcarpa, accounting for 56.43% of the total number of ancient trees, which may be related to the unique climatic conditions and historical cultivation habits of this area. The plant community was dominated by species from tropical Asia, accounting for 78.69% of the total flora, indicating that the area belonged to the tropical or subtropical climate zone, which is conducive to the growth of tropical plants. 2) Spatial distribution pattern of ancient trees: The spatial distribution of ancient trees in the study area showed obvious clustering characteristics, and the Moran’s I index was 0.903, indicating that their distribution had an obvious spatial correlation rather than a random distribution. The hotspots were mainly concentrated in Zuntan and Xinpo Towns, and near the junction of Laocheng and Shishan Towns. These areas may have had good natural conditions or may have been well-preserved historically. Cold spots were mainly concentrated in Haixiu, Changliu and Jinjiang towns. The number of ancient trees in these areas was relatively small, which may be related to the high intensity of human activities and frequent land development, which led to a reduction in the number of ancient tree resources. 3) Factors affecting the spatial distribution of ancient trees: The spatial distribution of ancient trees is affected by both natural and anthropogenic factors. Soil texture was the main natural factor affecting the spatial distribution of ancient trees, explaining 31.7% of the total variation in spatial distribution. The water retention capacity, drainage capacity, and nutrient content of the soil are crucial for the growth of ancient trees. Other important factors included rainfall (18.66%), soil organic matter content (18.48%), and population density (18.3%). According to the GWR model analysis, soil texture had an overall inhibitory effect on the spatial distribution of ancient trees, which may have been due to the fact that certain soil types restricted the growth of ancient trees. Precipitation, land-use intensity, and population density were the major factors. In areas with high population densities, human activities may have driven the expansion of the distribution area of ancient trees, thereby promoting their growth of ancient trees. This study showed that the distribution and quantity of ancient tree resources were affected by multiple factors, especially the combined effects of natural conditions and human activities. By comprehensively analyzing these factors, this study provides valuable insights into the spatial distribution characteristics of ancient tree resources and offers theoretical support and a scientific basis for the sustainable utilization and protection planning of ancient tree resources in volcanic lava areas.

Key words: ancient tree resources, spatial distribution patterns, Geographically Weighted Regression (GWR), volcanic lava landforms, northern Hainan Island

摘要：

古树承载着城乡的特色风貌和历史变迁，是不可再生的绿色文化遗产，也是推进乡村振兴与村落景观建设的重要生态资源。琼北火山熔岩区因其独特的地貌和土壤特征，孕育了丰富且具有显著特征的古树群落。基于该区域的古树资源普查数据，运用地理信息系统（GIS）和地理加权回归（GWR）模型，对古树空间分布格局及其影响因素进行了定量分析。结果表明，1）研究区古树资源丰富，共有古树521株，隶属8科12属15种。其中，树龄100－300 a的古树占比达95.59%；树种以桑科植物为主，其中榕树（Ficus microcarpa）数量最多，占56.43%；植物区系以热带亚洲区系植物为主，占总数的78.69%。2）研究区古树在空间分布上具有显著的聚集性和空间自相关性，Moran’s I指数为0.903；热点区域主要分布在遵谭镇、新坡镇周边区域以及老城镇和石山镇交界地带，而冷点区域则集中在海秀镇、长流镇以及金江镇一带。3）研究区古树的空间分布格局受到自然和人为因素的综合影响，其中土壤质地是影响研究区古树空间分布的主要因素，解释力为0.317，其次是降雨量（0.187）、土壤有机质含量（0.185）和人口密度（0.183）。通过GWR模型分析，土壤质地对古树空间分布总体表现为抑制作用；而降雨量、土地利用强度指数和人口密度以促进作用为主。研究结果可为火山熔岩区古树资源的可持续利用和保护规划提供理论支持与科学依据。

关键词: 古树资源, 空间分布, 地理加权回归（GWR）, 火山熔岩地貌, 海南岛北部