李昆,王玲,孙伟,王祥荣,李兆华.城市化下景观格局对河流水质变化的空间尺度效应分析[J].环境科学学报,2020,40(1):343-352
城市化下景观格局对河流水质变化的空间尺度效应分析
- Spatial effect of landscape pattern on river water quality under urbanization
- 基金项目：国家重点研发计划（No.2016YFC0502700）
- 孙伟
- 复旦大学环境科学与工程系, 上海 200433
- 王祥荣
- 复旦大学环境科学与工程系, 上海 200433
- 摘要：研究不同空间尺度上景观组成与结构变量对城市河流水质的影响程度和机制对于流域生态保护具有重要意义.以汉江流域襄阳城区段8个水质监测断面为中心，利用GIS软件以河流为对象生成8种空间尺度的缓冲区域，采用景观格局指数、冗余分析等方法与模型，识别景观格局对河流水质影响最有效的河岸带空间尺度，以及其与水质的相互关系.结果表明：①总体上农用地和城市用地的面积占比随缓冲区宽度上升而增加，而其他景观类型呈相反的变化特征.②景观格局在不同宽度缓冲区内对河流水质的影响具有空间尺度性，在300 m宽度解释能力最大，达到了78.5%.③景观组成变量中仅园地的面积比例（PLANDGAR）在400 m宽度上与水质指标存在相关关系.④影响水质变化的主要景观类型在小尺度上是城市用地，相对较大尺度上是林业用地，并且城市用地和林业用地的斑块密度（PDURB和PDFOR）对河流水质的负面影响最为显著.结果反映了300 m宽度是景观格局对水质变化的最佳临界区，景观组成变量对水质的影响程度小于景观结构变量.同时，不同景观类型对水质的影响机制也具有差异性，林业用地分布越密集、面积越大对进入水体的污染物净化作用越明显，而分散分布的城市区域对水质退化的作用显著.园地和草地因种植方式的差异对水质变化有一定的影响；而农业用地与水质之间没有发现显著的相关性.研究结果可以从宏观尺度上为流域水环境保护、景观优化与管理提供科学依据.
- Abstract：To figure out the effect of landscape composition and configuration on water quality in different spatial scales is important for ecological protection. In this study, the urban section of Xiangyang City of the Hanjiang River was selected as a case study. Eight water sampling sites in the study area were selected to analyze the variation of water quality. The land use/cover data and landscape metrics were analyzed at eight buffer widths by using ArcGIS and Fragstats software, respectively. The most effective spatial scale and the linkage between landscape patterns and water quality were identified through the redundancy analysis. The results show that:①the proportion of agricultural and urban land increased with the increase of buffer width, while other landscape types were opposite. ②The effect of landscape patterns on water quality was scale dependent at multiple buffer widths, and the total explanatory power was highest at the 300 m buffer width, accounting for 78.5%. ③Only the PLANDGAR was positively correlated with water quality at the 400 m buffer width among the landscape composition. ④Urban land was the main influence factor at smaller buffer widths, while forest land was the most important factor at wider buffer widths. The patch density of urban and forest land (PDURB AND PDFOR) had the most significant negative impact on the water quality. The results indicate that the strongest correlation between landscape patterns and water quality occurred at the 300 m buffer width, and the impact of landscape configuration on the water quality was more than that of landscape composition. Moreover, the influence mechanism of different landscape types on water quality was varied. Forest land with higher density and larger area had more obvious purification effect on water pollution, while the dispersed urban areas had a significant effect on water quality degradation. The difference of planting methods between garden plots and grass lands had some influence on water quality. However, no significant correlation was found between agricultural land and water quality. These findings could provide important scientific basis for water quality protection, landscape optimization and management in river basin.
