Macroinvertebrate and fish metrics are an essential tool in freshwater biomonitoring. Yet, many well-known stream metrics do not perform well as indicators of anthropogenic stress in lakes and wetlands, and there is a need to better understand the role of potential confounding factors such as latitude or environmental variables acting at regional scales. We attempt to untangle the relationships among metrics and watershed stressors while accounting for potential confounding factors using hierarchical partitioning, and test metric consistency across time, focusing on the Laurentian Great Lakes wetland monitoring programs conducted 10 years apart. Our results show that many frequently used metrics have high temporal variability and are significantly affected by spatial factors, most notably richness-type metrics used in numerous regional monitoring studies. Only a few metrics (invasive fish species richness and relative abundance, mayfly family richness and the relative abundance of Ephemeroptera, Trichoptera, Sphaeriidae and Odonata) had slightly better and more consistent correlations with watershed land use. We suggest that explicit consideration of confounding factors is essential in the context of large-scale monitoring programs, and focus on the less general (such as habitat-specific) metrics may be a more promising approach.

大型无脊椎动物和鱼类指标是淡水生物监测中必不可少的工具。但是，许多众所周知的流量指标不能很好地反映人为压力在湖泊和湿地中的表现，因此有必要更好地了解潜在混杂因素（例如纬度或环境变量在区域尺度上的作用）的作用。我们尝试理清度量标准和分水岭压力源之间的关系，同时使用分层划分来考虑潜在的混杂因素，并跨时间测试度量标准的一致性，重点是间隔10年进行的Laurentian Great Lakes湿地监测计划。我们的结果表明，许多常用的指标具有较高的时间变异性，并且受空间因素的影响很大，最值得注意的是，在众多区域监测研究中使用的丰富度指标。只有少数指标（入侵鱼类的物种丰富度和相对丰度，may蝇家族的丰富度以及phe，Trichoptera，Sphaeriidae和Odonata的相对丰度）与流域土地利用之间的关系更好，更一致。我们建议在大型监视程序的背景下，明确考虑混杂因素是必不可少的，并且关注较不通用的（例如特定于栖息地的）度量标准可能是一种更有希望的方法。