中文摘要：

溪流底栖动物群落结构受不同空间尺度环境因子的共同作用。基于2010年钱塘江中游流域60个样点的大型底栖无脊椎动物和环境变量数据,寻找与研究流域底栖动物群落结构变化密切相关的关键环境变量,解析流域尺度和河段尺度的环境因子对底栖动物群落的相对影响。PCA分析表明该区域的主要环境梯度是流域内的土地利用类型及其引起的溪流物理生境退化程度和水体营养状态。CCA分析发现影响底栖动物群落的流域尺度的关键环境变量是纬度、海拔、样点所在流域大小、森林用地百分比,河段尺度是总氮、总磷、钙浓度、二氧化硅浓度和平均底质得分。偏CCA分析得到两种尺度环境因子对底栖动物变异的总解释量为26.4%,流域尺度和河段尺度变量分别为总解释量的50%和31%;方差分解结果表明研究区域大型底栖无脊椎动物受到两种尺度环境因子的综合影响,且流域尺度环境因子较河段尺度环境因子更为重要,体现了其在溪流生态系统保护、恢复、监测和评价中的重要参考价值。

英文摘要：

Benthic macroinvertebrates are common,long-lived,sensitive to disturbance,and cost-effective to sample,which make them ideal biological indicators of aquatic system degradation,and they are used in stream biomonitoring worldwide.As biological indicators,macroinvertebrates can provide insight into the current and past conditions of waterbodies and they integrate the effects of cumulative stressors.However,patterns in stream macroinvertebrate assemblages are the result of a combination of processes acting at different spatial scales.Understanding the relative influence of environmental variables at different spatial scales can increase our ability to detect anthropogenic influences on stream integrity,as well as to assess and manage aquatic resources.In this study,we used Qiantang River basin as an example of a human disturbed watershed.Based on environmental and biotic data of 60 sites located in the middle section of Qiantang River basin,the specific aims of our study were：（1） to analyze the stream macroinvertebrate assemblages;（2） to identify the key environmental factors that are linked to variation in macroinvertebrate assemblages;and（3） to partition the additive effects of watershed-scale and reach-scale variables,as well as their interaction,on macroinvertebrate community composition.Environmental variables mainly included land use（urban,agriculture,forest,and the total impervious area）,geographical location,elevation,slope,stream order,and the area of sub-basin at the watershed scale,physical habitat conditions and water chemistry at the reach scale.Macroinvertebrate responses were characterized by the relative abundance of 262 taxa.Principal components analysis was used to explore the major environmental gradients,and canonical correspondence analysis was adopted to determine the relationships between environmental variables and macroinvertebrate composition.Variation partitioning was performed using partial canonical correspondence analysis（pCCA） to understand the relative importance