• 关键词：O/E指数  RIVPACS  随机森林  大型底栖无脊椎动物  生物评价  淮河

• 基金项目：国家水体污染控制与治理科技重大专项（No.2014ZX07204-006-01）；水利部公益性行业科研专项（No.201501007）；水利部交通运输部国家能源局南京水利科学研究院博士后基金（No.BH91401）

• 陈凯
• 1. 南京水利科学研究院生态环境研究中心, 南京 210024;2. 南京农业大学昆虫系水生昆虫与溪流生态实验室, 南京 210095

• 刘祥
• 南京水利科学研究院生态环境研究中心, 南京 210024

• 陈求稳
• 南京水利科学研究院生态环境研究中心, 南京 210024

• 王敏
• 南京水利科学研究院生态环境研究中心, 南京 210024

• 王丽
• 南京水利科学研究院生态环境研究中心, 南京 210024

• 王备新
• 南京农业大学昆虫系水生昆虫与溪流生态实验室, 南京 210095

• 摘要：本研究尝试采用我国已有的季风气候区底栖动物观测值（O）/期望值（E）比值模型，在无足够数量参照样点的情况下，建立淮河流域O/E指数健康评价指标体系，评价淮河流域典型水体底栖动物物种组成完整性现状。于2014年夏季（8月）和秋季（11月）分别调查了淮河流域20和25个典型水体的底栖动物和水质指标。O/E模型评价结果表明，监测点位在PC（Probabilities of Capture）>0和PC ≥0.5条件下，夏季和秋季的物种期望丰富度分别约为25和9.所有25个点位中，仅1个点位为健康，其余为一般或较差；模型控制自然梯度后O/E₀和O/E₅₀值均没有显著的季节性差异（p=0.565和0.229）。环境胁迫因子（TN、EC、COD_Cr和pH）和土地覆盖数据（水体比例、湿地比例、裸地比例、森林比例和草地比例）对秋季O/E₅₀和O/E₅₀-null的解释量高于夏季，TN是能够解释淮河流域典型水体夏季O/E指数变异最多的环境因子，pH和COD_Cr是能够解释秋季O/E指数变异最多的环境因子。研究表明，在缺少有效参照点位构建评价指标体系的情况下，在淮河流域应用已经构建的季风气候区底栖动物O/E指数模型进行健康评价是比较可靠的方法。

• Abstract：In the absence of a sufficient number of high quality reference sites, we attempted to use the well-developed macroinvertebrate-based Observed/Expected (O/E) models of Chinese monsoonal streams and rivers to calculate O/E index and assess the integrity of macroinvertebrate assemblages in the Huai River basin. We collected macroinvertebrate and water samples from 20 and 25 sites in summer (August) and autumn (November) 2014, respectively. We found that the expected taxa richness were 25 and 9 with PC ≥0 and PC≥0.5, respectively, insignificantly related to the season. Modeled O/E values with natural variation adjustment were higher than their related null O/E counterparts with no natural variation adjustment. Neither O/E₀ nor O/E₅₀ showed significant variation between seasons (p=0.565 and 0.229, respectively). The variation of modeled and null O/E₅₀ in autumn explained by environmental stressors (e.g., TN, EC, COD_Cr and pH) and land cover (e.g.,water bodies,wetland,bareland,forest, and grassland) variables were higher than those in summer. TN was the most selected variable that can explain variation of O/E values in summer; however pH and COD_Cr were the most selected variables in autumn. The health conditions of integrity of macroinvertebrate assemblage of all sites were in fair or poor class, except there was only one site in health class. Our results indicated that applying the well-developed macroinvertebrate-based O/E models of Chinese monsoonal streams and rivers is one relative robust method to assess the integrity of macroinvertebrate assemblages in the Huai River.

• Key words：O/E index  RIVPACS  random forest  macroinvertebrate  bioassessment  Huai River

• 
  摘要点击次数： 1451 全文下载次数： 1832