不同空间划分方式下袁河流域景观结构对水质的影响

Impacts of landscape structure on water quality under different spatial scales in the Yuan River Basin

近年来景观结构的水文效应受到关注,研究不同空间划分方式下景观结构对河流营养盐、重金属变化的影响机制对于流域生态保护具有重要意义.本文于2018年7月和2019年1月在袁河干流及支流38个采样点采集水样,测定水体营养盐类污染指标(DO、NO-3-N、NH+4-N、TP和DOC浓度)和重金属类污染指标(Cr、Mn、Fe、Cu、Zn、As、Cd和Pb浓度).基于前向变量选择、冗余分析等方法,筛选景观结构指标,量化其在不同尺度下对水体营养盐、重金属变化的影响.结果表明:(1)农田、斑块密度(PD)、斑块聚集指数(COHESION)及散布与并列指数(IJI)是影响水体营养盐变化的主要指标.林地、建设用地(Res)、平均最近邻体距离(ENNMN)和最大斑块指数(LPI)的组合是影响重金属变化的主要指标.(2)在河岸缓冲带与圆形缓冲区划分方式下,景观结构均在100 m尺度对营养盐变化解释能力最强,平均解释率分别为31.5%、24.3%,均在1000 m尺度对重金属变化解释能力最强,平均解释率分别为32.0%、42.6%.(3)100 m河岸缓冲带和子流域尺度分别是景观结构影响水体营养盐、重金属变化的最佳空间尺度,平均解释率分别为31.5%、42.8%.以上结果表明,针对水体不同的污染类型,采用对应的划分方式及缓冲尺度有助于提高定量分析精度,为流域水环境保护、景观优化与管理提供科学依据.

Hydrological effects of landscape structures have attracted great attention in recent years. Understanding the effect of landscape structures on nutrient and heavy metals in different spatial division scales is important for ecological protection. Here, we determined water quality parameters including nutrients(DO, NO-3-N, NH+4-N, TP and DOC) and heavy metals(Cr, Mn, Fe, Cu, Zn, As, Cd and Pb) from 38 sample sites in August 2018 and January 2019. Impacts of landscape structures on nutrients and heavy metals were quantified at different spatial scales based on forward variable selection and redundancy analysis. The results showed that:(1) Farmland, Patch density(PD), Patch cohesion index(COHESION) and Interspersion Juxtaposition Index(IJI) were key parameters affecting the variations of water nutrients. Forestland, Residential, Mean Euclidean nearest neighbor index(ENNMN) and Largest patch index(LPI) were key parameters affecting the variations of heavy metals.(2) For the riparian buffer zone and the circular buffer zone, the landscape structure explained the highest variation of nutritive salts at the 100 m scale, and the riparian buffer zone(31.5%) explained better than the circular buffer zone(24.3%). The landscape structure explained the highest variation of heavy metals at the 1000 m scale, and the riparian buffer zone(32.0%) explained worse than the circular buffer zone(42.6%).(3) The optimal spatial scales of landscape structure explaining the variations of nutritive salts was the 100 m riparian buffer zone(31.5%), and the optimal spatial scales of landscape structure explaining the variations of heavy metals was the sub-basin scales(42.8%). The above results indicate that using different division methods and buffer scales is helpful to enhance the accuracy of quantitative analysis, and provides a scientific basis for watershed environmental protection.