秦岭南麓小流域不同空间尺度景观格局对水质的影响分析

Influence of different spatial scale landscape patterns on water quality in a small watershed at the southern foot of the Qinling Mountains

了解江河流域源头区景观与河流水质之间的关系,有助于进行适当的景观规划以保护水源涵养区水质.本研究基于2019—2022年秦岭南麓丹江源头区的闵家河流域11个采样点的水质数据,采用冗余分析(RDA)、非参数偏差减小和1000次bootstrap模拟法,探讨了不同尺度景观特征对水质的影响,分析了沿景观梯度水质的突变情况,并确定了关键景观阈值范围.结果表明:(1)水质指标时空变化显著,非汛期总氮(TN)和总磷(TP)平均浓度较汛期分别降低了21%和14%,上游水质优于中下游水质.不同尺度的景观格局差异显著,河岸带相较子流域尺度的景观类型更丰富,空间异质性更强,破碎程度更高;(2)土地利用程度(L)、草地、农地、建设用地、破碎程度和空间异质性高的景观对水质有负效应,高连通度的优势林地景观对水质有明显的正效应;(3)景观格局对水质的影响具有空间尺度效应和季节差异,子流域尺度比缓冲区尺度的景观特征更能解释水质变化,且对汛期水质贡献更高;(4)子流域关键景观指标的阈值分别为:最大斑块指数(LPI)>72%、农地比例<11%、L<214、香浓多样性指数(SHDI)<0.7、斑块密度(PD)<48 n·km^(-2),一旦超过以上景观阈值,河流中TN浓度突变的累积概率将超过87.5%.本研究结果可为通过管理和优化江河源头区的景观特征来控制非点源污染提供参考.

Understanding the relationship between landscape and river water quality in the river headwaters area is helpful for appropriate landscape planning to protect water quality in water conservation areas.Based on the water quality data from 11 sampling points in the Minjiahe watershed in the headwaters area of the Dan River at the southern foot of the Qinling Mountains from 2019 to 2022.Redundancy analysis(RDA),nonparametric deviation reduction and 1000 bootstrap simulation method were used to explore the effects of landscapes at different spatial scales on water quality,analyze the abrupt change in water quality along the landscape gradient,and identify the key landscape threshold range.The results showed that:①The spatial and temporal variations in water quality were remarkable.The mean concentrations of total nitrogen(TN)and total phosphorus(TP)decreased by 21%and 14%in the non-flood season compared with those in the flood season,respectively.Water quality in the upstream was better than those in the middle and downstream.The landscape patterns at different scales were significantly different.Compared with sub-watershed scale,buffer scale has more abundant landscape types,stronger spatial heterogeneity and higher degree of fragmentation.②Land use degree(L),grassland,farmland,urban land,and landscapes with high fragmentation and spatial heterogeneity had negative effects on water quality,while advantageous forest landscapes with high connectivity had positive effects on water quality.③The impact of the landscape patterns on water quality had spatial scale and seasonal differences.Landscape characteristics at the sub-catchment scale explain water quality changes better than those at the buffer scale,and this phenomenon was more obvious during the flood season than during the non-flood season.④The thresholds for key landscapes indicators in the subcatchment were:largest patch index(LPI)>72%,the proportion of farmland<11%,L<214,Shannon's diversity index(SHDI)<0.7,patch density(PD)<48 n·km^(-2).Once these threshold ranges are exceeded,the cumulative probability of abrupt changes in TN concentration in the river will exceed 87.5%.The results are useful for controlling non-point source pollution by managing and optimizing the landscape characteristics of river headwaters areas.