Prediction of soil organic matter variability associated with different land use types in mountainous landscape in southwestern Yunnan province, China

•The spatial variability of SOM under different land use types was investigated.•Multiple linear regression models were built to estimate SOM content.•IDW and UK were employed to predict the spatial distribution pattern of SOM.•The RMSE of the established IDW model was a little lower than that of the UK model.

SOM is a crucial factor that indicates soil fertility and vegetation status and, to a considerable extent, influences the CO2 concentration in the atmosphere and even the global carbon cycle. We collected 294 SOM data from the secondary soil survey of Yunnan province in 1979 and investigated 210 soil sample sites (0–20 cm depth) in 2011 to enhance the understanding of the spatial variability of SOM between two phase data and its dominant influencing factors in the mountainous region in southwestern China. We examined whether land use types (farmland, grassland, forestland and scrubland), topographic conditions (elevation, slope and CTI) and vegetation coverage (NDVI) affect the spatial distribution and content of SOM in a mountainous region in southwestern China. The results indicated that, SOM content decreased among different land use types in the following order: forestland > scrubland > grassland > farmland. An accurate spatial prediction of SOM content has great significance in the estimation of the SOC pool. This study exhibited that the combined use of vegetation index and terrain attributes would result in a suitable method of predicting SOM distribution even in complex terrain. The prediction of spatial variability in SOM contents was achieved establishing IDW and UK. The root mean square error RMSE and ME methods were selected as comparison criteria to indicate prediction accuracy. Owing to the heterogeneous natural environment, the application of the UK and IDW methods is limited. And it is still challenging that further work was necessary to accurately predict spatial distribution of SOM by choosing appropriate methods in these mountain areas. This study laid a foundation to estimate and evaluate SOM sequestration for regional land use management.

Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slide