U.S. Geological Survey, North Carolina Water Science Center, 3916 Sunset Ridge Road, Raleigh, North Carolina 27607, USA.
Ecol Appl. 2010 Jul;20(5):1384-401. doi: 10.1890/08-1311.1.
Responses of benthic macroinvertebrates along gradients of urban intensity were investigated in nine metropolitan areas across the United States. Invertebrate assemblages in metropolitan areas where forests or shrublands were being converted to urban land were strongly related to urban intensity. In metropolitan areas where agriculture and grazing lands were being converted to urban land, invertebrate assemblages showed much weaker or nonsignificant relations with urban intensity because sites with low urban intensity were already degraded by agriculture. Ordination scores, the number of EPT taxa, and the mean pollution-tolerance value of organisms at a site were the best indicators of changes in assemblage condition. Diversity indices, functional groups, behavior, and dominance metrics were not good indicators of urbanization. Richness metrics were better indicators of urban effects than were abundance metrics, and qualitative samples collected from multiple habitats gave similar results to those of single habitat quantitative samples (riffles or woody snags) in all metropolitan areas. Changes in urban intensity were strongly correlated with a set of landscape variables that was consistent across all metropolitan areas. In contrast, the instream environmental variables that were strongly correlated with urbanization and invertebrate responses varied among metropolitan areas. The natural environmental setting determined the biological, chemical, and physical instream conditions upon which urbanization acts and dictated the differences in responses to urbanization among metropolitan areas. Threshold analysis showed little evidence for an initial period of resistance to urbanization. Instead, assemblages were degraded at very low levels of urbanization, and response rates were either similar across the gradient or higher at low levels of urbanization. Levels of impervious cover that have been suggested as protective of streams (5-10%) were associated with significant assemblage degradation and were not protective.
本研究在美国九个主要城市地区调查了底栖大型无脊椎动物对城市强度梯度的响应。在森林或灌木地被转换为城市用地的城市地区，无脊椎动物组合与城市强度密切相关。在农业和放牧地被转换为城市用地的城市地区，无脊椎动物组合与城市强度的关系较弱或不显著，因为低强度城市地区已经受到农业的退化影响。排序得分、EPT 类群数量和物种的平均耐污值是群落状况变化的最佳指标。多样性指数、功能群、行为和优势度指标不是城市化的良好指标。丰富度指标比丰度指标更能指示城市效应，而且来自多个栖息地的定性样本与所有城市地区的单一栖息地定量样本（急流或木质树桩）的结果相似。城市强度的变化与一组景观变量密切相关，这些变量在所有城市地区都是一致的。相比之下，与城市化和无脊椎动物响应强烈相关的溪流内环境变量在不同城市地区之间存在差异。自然环境决定了城市化作用的溪流生物、化学和物理条件，并决定了不同城市地区对城市化的响应差异。阈值分析几乎没有证据表明存在对城市化的初始抵抗期。相反，在城市化的极低水平下，群落就已经退化，而且响应率要么在整个梯度上相似，要么在低城市化水平下更高。已经提出的作为溪流保护的不透水覆盖水平（5-10%）与显著的群落退化有关，而且没有保护作用。
