Megan Mehaffey H, Nash Maliha S, Neale Anne C, Pitchford Ann M
U.S. Environmental Protection Agency, Washington, USA.
Environ Monit Assess. 2007 Jan;124(1-3):141-56. doi: 10.1007/s10661-006-9214-9. Epub 2006 Aug 1.
The objective of this study was to assess the applicability of using landscape variables in conjunction with water quality and benthic data to efficiently estimate stream condition of select headwater streams in the Mid-Atlantic Coastal Plains. Eighty-two streams with riffle sites were selected from eight-two independent watersheds across the region for sampling and analyses. Clustering of the watersheds by landscape resulted in three distinct groups (forest, crop, and urban) which coincided with watersheds dominant land cover or use. We used non-parametric analyses to test differences in benthos and water chemistry between groups, and used regression analyses to evaluate responses of benthic communities to water chemistry within each of the landscape groups. We found that typical water chemistry measures associated with urban runoff such as specific conductance and dissolved chloride were significantly higher in the urban group. In the crop group, we found variables commonly associated with farming such as nutrients and pesticides significantly greater than in the other two groups. Regression analyses demonstrated that the numbers of tolerant and facultative macroinvertebrates increased significantly in forested watersheds with small shifts in pollutants, while in human use dominated watersheds the intolerant macroinvertebrates were more sensitive to shifts in chemicals present at lower concentrations. The results from this study suggest that landscape based clustering can be used to link upstream landscape characteristics, water chemistry and biotic integrity in order to assess stream condition and likely cause of degradation without the use of reference sites. Notice: Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy.
本研究的目的是评估结合景观变量与水质及底栖生物数据，以有效估算大西洋中部沿海平原特定源头溪流的溪流状况的适用性。从该地区82个独立流域中选取了82条有浅滩区域的溪流进行采样和分析。根据景观对流域进行聚类，形成了三个不同的组（森林、农作物和城市），这与流域的主要土地覆盖或用途相吻合。我们使用非参数分析来检验各组之间底栖生物和水化学的差异，并使用回归分析来评估每个景观组内底栖生物群落对水化学的响应。我们发现，与城市径流相关的典型水化学指标，如电导率和溶解氯，在城市组中显著更高。在农作物组中，我们发现与农业相关的变量，如营养物质和农药，显著高于其他两组。回归分析表明，在污染物变化较小的森林流域中，耐污和兼性大型无脊椎动物的数量显著增加，而在以人类活动为主的流域中，不耐污的大型无脊椎动物对较低浓度化学物质的变化更为敏感。本研究结果表明，基于景观的聚类可用于将上游景观特征、水化学和生物完整性联系起来，以便在不使用参考站点的情况下评估溪流状况及可能的退化原因。注意：尽管本工作经过美国环境保护局审查并批准发表，但它不一定反映该机构的官方政策。
