中文摘要：

引入地方特质作为外生变量,构建旅游者地方依恋与环保行为倾向的结构模型,运用验证性因子分析与结构方程模型方法,以世界自然遗产地九寨沟为对象,探讨了自然观光地地方特质与旅游者地方依恋和环保行为倾向的内在关系.结果表明：①地方依恋的不同维度对不同类型的环保行为倾向的影响存在显著差异;②地方特质对旅游者地方依恋具有显著的正向影响,表明九寨沟地方特质是旅游者地方依恋构建的重要因素;③地方特质通过地方依恋对旅游者环保行为倾向产生显著的间接影响,是解释九寨沟旅游者环保行为倾向的有效因子,也是旅游者地方依恋与环境行为倾向关系中重要的影响因素;④九寨沟地方特质不同维度的影响程度存在差异,显著的地方特质（自然环境属性）具有更明显的影响效应.

英文摘要：

Along with the increasing concern about the relationship between human and environment,the relationship between tourist place attachment and their pro-environment behavioral intentions has been studied by a few researchers.However,＂place＂ has always been taken as the context without an explicit explanation of the connotation and its potential role.The specific attributes of a place,referring to the comparative strength or uniqueness of the place,have been hardly considered in the research framework of pro-environment behaviors and intentions although it is the fundamental and indispensable element of the human-environment interactions,especially in a distinctive and impressive destination such as Jiuzhaigou.Jiuzhaigou is a natural World Heritage site in China,famous for its splendid natural scenery and the Tibetan customs and traditions.Using Jiuzhaigou as the study case,this article took the specific attributes of the place（containing natural environment and social environment of Jiuzhaigou perceived by tourists） as the antecedents,place attachment（containing place dependence and place identity） as the intermediaries,and pro-environment behavioral intentions（containing low effort pro-environment behavioral intentions and high effort pro-environment behavioral intentions） as consequences to establish a structural model for further exploration of the relationship between tourists＇ place attachment and their pro-environment behavioral intention,and for an experimental examination of the role of specific attributes of a place in that relationship.Data were collected by structured questionnaire survey carefully designed and amended.One variable was removed after the initial reliability and validity analysis.After confirmatory factor analysis,the model was tested.The overall fit of the measurement model suggested a reasonable model fit to the data given the sample size and number of indicators.Finally two of eight hypotheses were rejected and all of the rest were accepted.Four main conclusions have b