在较大的空间尺度上生态位模型是预测物种潜在分布的有效途径之一。为了探讨在热带天然林景观中木本植物(限于乔木和灌木)主要关键种的潜在分布,在对海南岛霸王岭的热带天然林进行按公里网格样方调查的基础上,采用演替地位和最大潜在高度两个功能性指标对物种进行了功能群划分,并在功能群框架下运用优势度指数法进行了关键种的确定;采用基于地理信息系统(Geographic information system, GIS)的基于规则集合预测的遗传算法(Algorithm for rule-set prediction, GARP)生态位模型对主要关键种的地理分布进行了预测,并应用受试者工作特征分析进行了模型精度验证;应用多元线性回归分析对影响各关键种潜在分布的关键因子进行了确定。结果表明:除了顶极次林层乔木功能群和顶极主林层乔木功能群外,在先锋种功能群、顶极灌木种功能群和顶极超冠层乔木功能群中采用优势度指数法划分出的关键种较为理想;一般来讲,在进行预测的8个关键种中,除了先锋主林层乔木种海南杨桐(Adinandra hainanensis),其它3个先锋种毛稔(Melastoma sanquiueum)、银柴(Aporosa chinensis)和枫香(Liquidambar formosana) 在研究区北部、西部以及西南部均具有较高的发生概率,而顶极种除了顶极超冠层乔木种南亚松(Pinus merkusii)外,九节(Psychotria rubra)、高脚罗伞(Ardisia quinquegona)和海南椎(Castanopsis hainanensis)具有相似的潜在分布格局,在研究区中部、东南部和南部地区具有较高的发生概率;相关分析表明极端最低温、年均温、极端最高温、年均降水量、海拔和坡向6大因子是影响研究区关键种潜在分布的关键因子;精度检验表明,GARP模型对8个关键种的潜在分布预测效果均较好,而其中又以银柴和海南椎的预测精度最高。

Aims Our major objectives were to 1) identify keystone species within the context of functional groups, 2) develop potential distributional predictions for keystone species using ecological niche model, 3) confirm factors determining potential distributions of keystone species, and 4) test if the performances of ecological niche model are better than those of a random model and differ in predicting different keystone species.

Methods Based on the investigation of 135 plots in a natural tropical forest landscape, we classified woody plant functional groups based on successional status and potential maximum height. Keystone species within each functional group were identified using a dominance index (DI). We used the genetic algorithm for rule-set prediction (GARP) to estimate the keystone species' potential distribution and then used the receiver operating characteristics to evaluate predictive performance. Applying multiple linear regression analysis, we identified major factors determining potential distributions of keystone species.

Important findings Identification of keystone species within pioneer species, climax shrub and emergent tree functional groups was clearer than within climax subcanopy and climax canopy tree functional groups. Generally, among the eight keystone species, pioneer species Melastoma sanquiueum, Aporosa chinensis and Liquidambar formosana (but not Adinandra hainanensis) have high probability of occurrence in the north, west and southwest regions of Bawangling. However, climax species Psychotria rubra, Ardisia quinquegona and Castanopsis hainanensis (but not Pinus merkusii) have high probability of occurrence in the central, southeast and south regions. Minimum and maximum temperature, mean annual temperature and precipitation, aspect and altitude were the key factors determining potential distributions of keystone species. Evaluation of GARP model's performance indicated excellent predictive ability of all eight keystone species' distribution. This study suggests the DI method is more suitable to identify keystone species within woody plant functional groups in which a single or a few species are dominant. Findings will assist decision makers in planning conservation and management policies in tropical rainforest areas.