Reflectance spectroscopy detects management and landscape differences in soil carbon and nitrogen.

Abstract

Many studies have calibrated visible and near-infrared (VNIR) diffuse reflectance spectroscopy (DRS) to various soil properties; however, few studies have used VNIR DRS to detect treatment differences in controlled experiments. Therefore, our objective was to investigate the ability of VNIR DRS to detect treatment differences in topsoil organic C (SOC) and total N (TN) compared with standard dry combustion analysis. A long-term (since 1991) experiment in central Missouri, where cropping systems were replicated across a typical claypan soil landscape was studied. Soil samples from two depths (0–5 and 5–15 cm) were obtained in 2008 at summit, backslope, and footslope positions for three grain cropping systems. Estimates of SOC by VNIR DRS using oven-dried soil samples and an independent calibration set were very good, with R2 = 0.87 and RMSE = 2.4 g kg−1. Estimates of TN were somewhat less accurate (R2 = 0.79, RMSE = 0.24 g kg−1). Field-moist VNIR DRS results were also good, but with 13 to 17% higher RMSE. Trends in differences among treatment means were very similar for dry combustion, oven-dry soil VNIR, and field-moist VNIR. Dry combustion was best at separating treatment means, followed by dry soil VNIR and fi eld-moist VNIR. Differences among methods were relatively minor for 0- to 5-cm depth samples but more pronounced for 5- to 15-cm samples. Efficiency of the VNIR method, particularly when applied ton field-moist soil, suggests that it deserves consideration as a tool for determining near-surface SOC and TN differences in field experiments.