The influence of physical habitat and land use on stream fish assemblages in southeastern Michigan
The influence of physical habitat and land use on stream fish assemblages in southeastern Michigan
复制标题
DOI：
--
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
Matthew J. Diana;J. Allan;D. Infante
中科院分区：
文献类型：
--
作者：
Matthew J. Diana;J. Allan;D. Infante
- 调查了休伦河和葡萄干河48个河段的土地利用和不稳定的物理生境对鱼类群落生物条件的影响。流域和100 m河流缓冲区内的农业和湿地数量与不稳定的物理生境关系最密切，这两类土地利用/覆盖之间呈负相关（r =-0.70，p = 0.01）。农业与高水平的沉积和降低流动稳定性，而湿地与低沉积和稳定的流量。生物完整性指数（IBI）与低沉积、稳定流动和细砾石（2-8 mm）的存在呈正相关。在缓冲区内，与农业用地的相关性不显著，与自然土地覆盖（森林+湿地）的相关性显著。最好的线性回归模型，使用物理栖息地和土地利用变量从所有网站充分预测IBI分数（调整R = 0.52）。然而，当休伦和葡萄干盆地分别处理，包括变量的一些不同，模型拟合增加（休伦调整R = 0.76，葡萄干调整R = 0.79），表明鱼类组合的关系，物理栖息地和土地利用不同的流域之间。葡萄干模型包括土地覆盖变量，而休伦模型只包括与物理栖息地有关的变量。因此，在这些盆地的instantums栖息地和土地覆盖可能发挥不同的作用，这表明形成单独的模型，个别盆地时，有足够的数据是有益的。
—The influence of land use and instream physical habitat on biotic condition of fish assemblages was investigated for 48 stream reaches in the Huron and Raisin rivers. The amount of agriculture and wetland in the catchment and 100-m stream buffers had the strongest relationships with instream physical habitat, and these two categories of land use/cover were negatively correlated with each other (r = –0.70, p = <0.01). Agriculture was associated with high levels of sedimentation and reduced flow stability, while wetland was associated with low sedimentation and stable flows. The index of biotic integrity (IBI) was positively related to low sedimentation, stable flows, and the presence of fine gravel (2–8 mm). It was not significantly correlated with agricultural land use, but was positively related to natural land cover (forest + wetland combined) in the buffer. The best linear regression model using physical habitat and land-use variables from all sites adequately predicted IBI scores (adjusted R = 0.52). However, when the Huron and Raisin basins were treated separately, some of the included variables differed, and model fit increased (Huron adjusted R = 0.76, Raisin adj. R = 0.79), indicating that relations of fish assemblages to physical habitat and land use differed between basins. The Raisin model included land cover variables, while the Huron model included only variables related to physical habitat. Thus instream habitat and land cover may play different roles in these basins, suggesting the benefit of forming separate models for individual basins when sufficient data are available.