Multiple landscape “source–sink” structures for the monitoring and management of non-point source organic carbon loss in a peri-urban watershed

We established successfully the Location-weighted landscape Contrast Index (LCI) with three factors of landscape pattern (land-use type, composition and configuration) in the view of the source–sink structures of three organic carbon format loss processes. To reveal the relations between soil organic carbon loss and landscape indicators, we should firstly quantify the variables that denote organic carbon loss status. The sediment delivery distributed model and the Pollution Load (PLOAD) model (R2>0.9; ENS>0.9) were integrated to simulate sediment yield, total organic carbon (TOC) loss, adsorbed organic carbon (AOC), and dissolved organic carbon (DOC). Results by Pearson correlation analysis indicate that the dominant factor in the integrated model is the rainfall factor explaining >80% of the variations in both sediment and organic carbon loss. Aside from precipitation, landscape pattern is also a principal factor that correlates with sediment yield and organic carbon loss. Landscape indicators, namely, LCITOC and LCIAOC, are significantly related to the sediment delivery ratio, and they explain 87% and 81% of the variation in sediment yield rate, respectively; however, they do not consider precipitation, with the rainfall factor dropped. LCITOC, LCIAOC, and LCIDOC explain 91%, 89%, and 87% of the variation in the corresponding organic carbon formats, respectively. The landscape indicators show that the landscape pattern is generally unfavorable to AOC loss in most sub-watersheds, whereas it is favorable to DOC and TOC format losses in the whole watershed. These differences among loss formats are considerably attributed to different sink–source compositional structures. Flow length derived from the paddies and rural settlements along river networks is the most unfavorable factor related to spatial configuration of organic carbon source–sink. In this regard, this study proposes the further optimization of the landscape pattern to increase or prevent the flow path of soil and organic carbon from source land-use types to river channels. This goal can be achieved through the development of protection forest belts in hilly areas and around agricultural lands.