@article{8667299,
  abstract     = {{The soil fertility index (SFI) is an indicator that is commonly used to evaluate the soil fertility in the rice paddy regions of China. However, calculating an SFI requires the laboratory measurement of multiple soil properties, which adds to the costs and complexity of conventional methods. Visible and near-infrared (vis-NIR, 400-2,500 nm) spectra might offer opportunities to cost-effectively and rapidly assess soil properties and the SFI. In this study, we evaluated soil fertility properties and their derived SFIs for paddy fields in southern China using vis-NIR spectra with the partial least-squares regression model. The results showed very good prediction accuracy of vis-NIR spectroscopy for soil organic matter, clay and sand contents, with the former two having an apparent spectral response in the near-infrared range. A good predictive ability was obtained for total nitrogen, silt, available nitrogen and total potassium. In contrast, the prediction accuracy was only moderate for cation exchange capacity and available phosphorus, and it was poor for the pH, total phosphorus and available potassium without apparent spectral responses. A bootstrapped partial least squares regression model predicted the SFI directly from the vis-NIR spectra accurately (R-2 of 0.80 and ratio of performance to interquartile range of 3.12), whereas SFI computed from vis-NIR estimates of the individual indicators was less accurate. This shows that vis-NIR spectroscopy can improve the efficiency of soil fertility assessments within the cultivated paddy rice regions of southern China. Highlights

Soil fertility index (SFI) can be predicted directly with spectroscopy. Prediction accuracy of vis-NIR spectroscopy was good for SOM, TN and clay. SFI estimated directly from vis-NIR spectra was superior to SFI calculated from predicted individual indicator. Successful prediction of SFI from vis-NIR spectra can be attributed to the high significance of SOM and TN.}},
  author       = {{Yang, Meihua and Mouazen, Abdul and Zhao, Xiaomin and Guo, Xi}},
  issn         = {{1351-0754}},
  journal      = {{EUROPEAN JOURNAL OF SOIL SCIENCE}},
  keywords     = {{Soil Science,paddy rice,partial least-squares regression (PLSR),soil fertility index,visible and near-infrared spectroscopy,LEAST-SQUARES REGRESSION,REFLECTANCE SPECTROSCOPY,QUALITY,PREDICTION,EVALUATE}},
  language     = {{eng}},
  number       = {{4}},
  pages        = {{615--626}},
  title        = {{Assessment of a soil fertility index using visible and near-infrared spectroscopy in the rice paddy region of southern China}},
  url          = {{http://doi.org/10.1111/ejss.12907}},
  volume       = {{71}},
  year         = {{2020}},
}

Altmetric

View in Altmetric

Web of Science

Times cited: