基于地理探测器和随机森林模型的区域景观游憩服务潜在供应评估方法研究——以亚热带湿润淮阳低山景观区为例  

Assessment method for potential supply of recreational services in regional landscapes based on Geographical Detector and Random Forest model

作　　者：李业涵 周媛[1] 吴雪飞[1] LI Yehan;ZHOU Yuan;WU Xuefei(College of Horticulture and Forestry Sciences of Huazhong Agricultural University,Wuhan 430070,China)

机构地区：[1]华中农业大学园艺林学学院,武汉430070

出　　处：《生态学报》2024年第13期5629-5645,共17页Acta Ecologica Sinica

基　　金：国家自然科学基金面上项目(32371945);国家重点研发计划(2019YFD1100401)。

摘　　要：游憩服务(RS)是自然对人类的贡献(NCP)之一,也是区域尺度下主要的人类需求,由于其具有无形、主观、非物质等特点,量化RS一直是当前研究的难点与热点。既往研究的评估方法依赖于利益相关者的认知,缺乏通用性,并多为对各类土地类型或植被的RS潜在供应能力进行“一刀切”式直接赋值,鲜有考虑地域特征。研究旨在探究景观特征及其要素如何影响RS供给,从而提出一种因地制宜的RS潜在供应量客观评估方法。以“亚热带湿润淮阳低山景观区”为例,首先将风景资源点分为5类,利用核密度分析量化各类RS实际供应;然后利用地理探测器分类评估景观特征的RS贡献度;再基于随机森林模型对景观特征要素与RS潜在供应能力进行关联量化分析;最后测算出各景观类型的RS潜在供应量。研究结果表明:(1)不同景观特征的RS贡献度具有差异,不同景观特征要素具备不同的RS潜在供应能力;(2)景观特征的RS贡献度和景观特征要素的RS潜在供应能力在不同地理气候条件下具有空间分异特征,应当因地制宜进行评估;(3)景观特征的贡献度和景观特征要素的RS潜在供应能力取决于RS的类型,在评估RS潜在供应量时,应对RS进行分类探讨;(4)由于景观类型是景观特征要素的聚类簇,基于景观类型测算RS潜在供应量可以反映出若干景观特征要素RS潜在供应能力的协同性。强调了景观特征及其要素在RS评估方面的关键作用,所提出的量化方法突破了主观性的局限,具有客观优势,最终可为全域旅游规划和风景资源点建设选址提供参考。Recreational services(RSs)are one of nature′s contributions to people and represent a primary human need at the regional scale.Due to their intangible,subjective,and non-material characteristics,quantifying RS has been a challenge and hot topic in current research.Previous research methods for evaluation have relied on stakeholders′perceptions,lacking universality,and often involved a one-size-fits-all direct assignment of RS potential supply capabilities to various types of land or vegetation,with little consideration for regional characteristics.This study aims to explore how landscape characters and their elements influence the supply of RS,thereby proposing an objective assessment method for the potential supply of RS that is tailored to local conditions.Taking the Subtropical Humid Huaiyang Low Mountain Landscape Region as an example,the study firstly categorizes scenic resource points into five types and quantifies the actual supply of each type of RS using kernel density analysis.Then,the Geographical Detector is employed to categorically assess the contribution of landscape character to RS.This is followed by a correlational quantitative analysis of landscape character elements and the potential supply capability of RS based on the Random Forest model.Finally,the potential supply of RS for each landscape type is calculated.The results show that:(1)The contribution of different landscape characters to RS varies,and different landscape character elements possess varying potential supply capabilities for RS;(2)The contribution of landscape character to RS and the potential supply capability of landscape character elements for RS exhibit spatial differentiation under different geographical and climatic conditions,and assessments should be tailored to local conditions;(3)The contribution of landscape character and the potential supply capability of landscape character elements for RS depend on the type of RS.When assessing the potential supply of RS,it is necessary to classify and discuss RS;(4)Since lands

关 键 词：游憩服务 区域景观 景观特征 景观特征要素 地理探测器 随机森林模型 

分 类 号：F592.7[经济管理—旅游管理]