Estimating a Soil Quality Index with VNIR Reflectance Spectroscopy

Sensor-based approaches to assessment and quantification of soil quality are important to facilitate cost-effective, site-specific soil management. The objective of this research was to evaluate the ability of visible, near-infrared (VNIR) diffuse reflectance spectroscopy to estimate multiple soil quality indicators (SQIs) and Soil Management Assessment Framework (SMAF) scores. A total of 234 soil samples from two depths (0-5 and 5-15 cm) were obtained in 2008 from 17 agricultural management systems located in the Central Claypan Region of Missouri, USA. The VNIR spectra were obtained on oven-dried and field-moist soil, and calibration models were developed with partial least squares (PLS) regression. Models were evaluated using the coefficient of determination (R-2), residual prediction deviation (RPD), and the ratio of prediction error to interquartile range (RPIQ). The most reliable estimation results were achieved using oven-dry soil for organic C, beta-glucosidase, total N, the biological SMAF score, the organic C score, and the beta-glucosidase score (R-2 >= 0.76, RPD >= 2.0, RPIQ >= 3.2). Using field-moist soils, the most reliable estimation results were achieved for organic C and the organic C score (R-2 >= 0.80, RPD >= 2.1, RPIQ >= 3.6). Incorporating the bulk density score and P score as auxiliary variables with the VNIR spectra improved estimation of the overall SMAF soil quality score for oven-dry soil (R-2 = 0.76, RPD = 2.0, RPIQ = 3.1) and field-moist soil (R-2 = 0.75, RPD = 1.9, RPIQ = 2.8). These results demonstrate the robustness of VNIR estimation of biological SQIs, and illustrate the potential for rapid, in-field quantification of soil quality by fusing VNIR sensors with auxiliary data obtained from complementary sensors or supplemental analyses.