Nonlinearity and threshold effects of landscape pattern on water quality in a rapidly urbanized headwater watershed in China

Investigating the relationship between landscape pattern and water quality is essential for making sustainable landscape development policies. A clear understanding of such relationship remains elusive partly due to ubiquitous nonlinearity, behind which the landscape threshold can potentially act as quantifiable target of landscape conservation. In the upper Huai River Basin in China, six landscape metrics were quantified for forested, agricultural, and residential land use categories at broad sub-basin and finer riparian buffer scales. Five water quality indicators were collected from 24 monitoring stations during the period of 2006-2012. On the basis of nonlinear regression models, including power, exponential, quadratic, and segmented regressions, oneto-one relations between landscape metrics and water quality variables were fitted with R2 generally varying between 0.32 and 0.74. Meanwhile, three types of defined landscape thresholds were also identified on the basis of the fitted nonlinear relations. From the perspective of spatial scale dependence, the percentage of landscape area and largest patch index could well explain water quality variables at a broad sub-basin scale; the opposite could be said for patch density. The type-1 and type-2 landscape thresholds can act as the long-term targets where landscape pattern management strategy should be timely adjusted. By contrast, the type-3 landscape threshold can be set as a short-term target of landscape conservation to ensure water quality up to a certain management requirement. This study suggests that considering nonlinearity and threshold effect could provide insights into the complicated landscape pattern-water quality relations and thus provide a scientific basis for balancing the coupled socioecological system of landscape development and water quality protection.

Automated summary

The study suggests that considering nonlinearity and threshold effect can provide insights into the complicated relationship between landscape pattern and water quality. Research conducted in the upper Huai River Basin in China found that landscape area percentage and largest patch index could effectively explain water quality variables, while patch density had a weaker effect. The study also identified different types of landscape thresholds and highlighted their importance as long-term or short-term targets for landscape conservation.