陷阱法和Winkler法调查土壤节肢动物多样性比较：以千岛湖岛屿封闭生境研究为例
A comparison of pitfall trapping and the Winkler method for investigating soil arthropod diversity: a case study on the
黄杰灵**,胡广2,袁金凤2,罗媛媛1
点击：2245次 下载：29次
DOI：
作者单位：1. 中国计量学院生命科学学院杭州310018；2. 浙江大学生命科学学院杭州310058
中文关键词：土壤节肢动物， 取样方法， 千岛湖， 陆桥岛屿， 研究尺度
英文关键词：soil arthropod, sampling methods, Thousand Island Lake, landbridge island, research scales
中文摘要：
陷阱法和Winkler法是调查土壤动物的两种常规方法，然而这两种方法的调查效率各有优劣。本研究于2010年秋季，在千岛湖中心湖区选取了15个面积不同的岛屿，同时采用Winkler法和陷阱法采集岛屿上的土壤节肢动物，在岛屿这一封闭生境中比较两种方法收集土壤节肢动物的效率。结果表明，两种方法捕获土壤动物类群丰富度差异不显著，但多样性指数差异极显著。Winkler法对常规土壤节肢动物类群的采集效率优于陷阱法，尤其对运动较缓慢、活动范围较小的土壤节肢动物类群的采集具有优势；陷阱法则更优于采集运动能力较强、活动范围较大的类群。基于样方的稀疏曲线结果说明，Winkler法能用少量样方快速获取研究区域的土壤节肢动物群落基本组成，推荐在面积较小的岛屿上使用；而在面积较大的岛屿上使用陷阱法能够获取更多的类群。Chao-Jaccard相似性系数比较则显示两种方法所取的土壤节肢动物相似性在大型岛屿上差异较大，说明大样本数据的采集需要两种方法同时使用能够提高数据的完整性和可靠性。本研究的结果为土壤节肢动物研究的方法选择提供了数据支持，具有一定的实际应用性。
英文摘要： The Winkler method and pitfall trapping are two conventional methods used for soil fauna surveys, with each having advantages at different aspects and scales. In autumn 2010, we select 15 different areas of the island, we compared the efficiencies of these methods for collecting soil arthropods on landbridge islands in the Thousand Island Lake (TIL), Zhejiang, China. The results show that, Taxon richness was not significantly different between the two methods but the diversity indices were significantly different. The Winkler method is superior to pitfall trapping in collecting soil arthropods, especially for those taxa with slower movement and smaller territories, whereas pitfall trapping is more inclined to collect more mobile species. The samplebased rarefaction curves showed that the Winkler method can quickly achieve the basic list of local soil arthropods with fewer samples, and is thus recommended for smaller islands, whereas pitfall traps can capture more taxa on the larger islands. Analysis of ChaoJaccard similarity coefficients showed significant differences between the community compositions estimated by the two methods on large islands. This suggests that large scale sampling requires the use of both methods to improve the integrity and reliability of data.