摘要:基于河流水质观测数据和Landsat 8影像数据，分析厦门后溪2013至2017年干季和湿季的水质时空变化特征，研究沿城乡梯度河流水质变化与流域景观特征的关系。结果表明：后溪上游（饮用水水源地水体）和下游（景观水体）溶解氧（DO）分别符合《地表水环境质量标准（GB3838-2002）》Ⅱ类标准和Ⅴ类标准，但是总氮（TN）和总磷（TP）有不同程度的超标，其中TN的超标比例较高。后溪上游土地利用类型中70%以上为林地，具有涵养水源、保护水质的作用；下游建设用地和耕地比重增加，TN、TP和叶绿素a（Chl a）显著高于上游。耕地和建设用地是提高TN、TP和Chl a等水质参数的主要土地利用类型，斑块密度、香农多样性指数和源汇景观指数与各水质指标也显著相关。冗余分析表明，流域景观特征可解释70%以上水质变化，对TN和TP的影响在湿季更大，而对Chl a的影响在干季较大。土地利用组成、配置、距离、高程和坡度对流域水质均有较大影响，其中土地利用组成对Chl a的影响较大，而景观指数则对TN和TP的影响较大。因此，在流域尺度加强土地利用的规划与综合管理，通过减少和控制地表径流面源污染的途径在一定程度上可降低人类活动对水质的不良影响。



Abstract:In the current context of urbanization and climate change, anthropogenic activities have changed land use and land cover, which seriously affect water quality and riverine ecosystems. The impact of land use changes on water quality has great temporal and spatial heterogeneity. Here, we studied the relationships between water quality and landscape characteristics of the Houxi River watershed covering 74% of surface area of the Jimei district in Xiamen City. We analyzed spatial and temporal changes in the water quality from 2013 to 2017 in both dry and wet seasons. Our results showed obvious spatial heterogeneity from the upstream to downstream in water quality. Dissolved oxygen (DO) in the Houxi River upstream and downstream river sections met categories Ⅱ and Ⅴ of surface water quality standards (GB3838-2002), respectively, whereas total nitrogen (TN) and total phosphorus (TP) exceeded the national standards in most cases. Shidou and Bantou reservoirs in the upstream of the watershed were mainly in the range of light to middle eutrophic conditions during the study period. The percent of forest land was higher than 70% in the upstream watershed, which could significantly protect water quality. The proportion of built land and cropland associated with patch density (PD), Shannon's diversity index (SHDI), and location-weighted landscape index (LWLI) increased from the upstream to downstream, leading to increasing concentrations of TN, TP, and Chl a. Statistical analyses, including correlation analysis, regression analysis, redundancy analysis, and a structural equation model showed that, cropland and built land were the dominant land use types that could potentially affect water quality in complex ways. The relationships were also strong between the landscape index and DO, TN, TP, and Chl-a. The RDA results explained 78.6% and 74.0% in water quality variation in the dry and wet seasons, respectively. The influence of landscape characteristics on TN and TP, and its influence on Chl a was more pronounced in the wet and dry seasons, respectively. Overall, our results showed that the land use patterns, patch density, relative distance, elevation, and slope degree had a great impact on water quality in the watershed. Therefore, it is necessary to adopt appropriate land use management strategies in the watershed, particularly by controlling the runoff linked with agriculture and urbanization.





PDF全文下载地址:

https://www.ecologica.cn/stxb/article/pdf/stxb201803280625