Effects of comprehensive landscape patterns on water quality and identification of key metrics thresholds causing its...

Comprehensive landscape patterns influence water quality with multiple factors, complex processes, and scale dependence. However, studies identifying landscape thresholds causing abrupt water quality changes and characterizing the contribution of topography to water quality are still limited. Exploring the impact mechanisms of natural geographical and landscape characteristics on spatial and seasonal water quality variations is conducive to watershed water resource protection and ecosystem restoration. Based on water quality monitoring data of Minjiahe River in the typical headwater area of the upstream Dan River in China from 2019 to 2021, we employed redundancy analysis, partial redundancy analysis, and nonparametric change-point analysis to analyze the relationship between stream water quality and multi-spatial scale comprehensive landscape patterns, to obtain the interactive and independent contributions of different landscape categories at multi-spatial scales on water quality, and to find the key landscape threshold leading to abrupt changes in water quality. Results showed that landscape configuration, landscape composition, and topographic factors collectively explain over 89.1% of water quality variation. Most seasonal variations in water quality were primarily caused by landscape configuration. The landscape composition was mainly responsible for the differences in water quality variations among spatial scales. The topographic factors made the least independent contribution and had a potential impact on overall water quality variation. In order to protect the water quality of streams, it is more reasonable to regulate the landscape at different scales. At the sub-catchment scale, interspersion and juxtaposition index (IJI) and landscape shape index (LSI) should be controlled below 82% and 22. At the 100 m riparian scale, farmland, urban land, IJI, and LSI should be controlled below 29%, 6.5%, 92%, and 26, respectively. Our results provide important guidance for optimizing landscape patterns and water conservation in the watershed.