The use of Vis‐NIR spectral reflectance for determining root density: evaluation of ryegrass roots in a glasshouse trial

Abstract

This paper reports the use of visible/near-infrared reflectance spectroscopy (Vis-NIRS) to predict pasture root density. A population of varying grass root densities was created by growing Moata ryegrass (Lolium multiflorum Lam.) for 72 days in pots of Ramiha silt loam (Allophanic) and Manawatu fine sandy loam (Recent Fluvial) (60 pots for each soil) differentially fertilized with nitrogen (N) and phosphorus (P) in a glass house experiment. At harvest, the reflectance spectra (350-2500 nm) from flat sectioned horizontal soil slices (1.3 cm depth), taken from 57 selected pots, were recorded using a portable spectroradiometer (ASD FieldSpec Pro, Boulder, CO). Root densities within each of the soil slices were measured using a wet sieving technique. A large variation in root densities (0.46-5.02 mg dry root cm À3 ) was obtained from the glass house experiment as plant growth responded to the different soils and rates of N and P fertilizer treatment. Pots of the Manawatu soil contained greater ryegrass root densities (1.76-5.02 mg dry root cm À3 ) than pots of the Ramiha soil (0.46-3.84 mg dry root cm À3 ). Each soil had visually distinct reflectance spectra in the range 470-2440 nm, but different root masses produced relatively small differences in reflectance spectra. The first two principal components (PC1 and PC2) of a principal component analysis of the first derivative of the spectral reflectance accounted for 71.3% of the spectral variance and clearly separated the Ramiha and Manawatu soils. PC1, which accounted for 58.4% of the spectral variance, was also well correlated to root density. Partial least squares regression (PLSR) of the first derivative of the 10 nm spaced spectral data against measured root densities produced calibration models that allowed quantitative estimates of root densities (without removing outlier, r 2 cross-validation ¼ 0.78, ratio of prediction to deviation (RPD) ¼ 2.14, root mean squares error of cross-validation (RMSECV) ¼ 0.60 mg cm À3 ; with removing outliers, r 2 cross-validation ¼ 0.85, RPD ¼ 2.63, RMSECV ¼ 0.47 mg cm À3 ). The study indicated that spectral reflectance measurement has the potential to quantify root density in soils.

Net changes of soil C stocks in two grassland soils 26 months after simulated pasture renovation including biochar addition

“…Four 5‐cm‐diameter, 2‐cm‐depth aluminium rings were randomly inserted into the soil surface of each core layer. Each ring was scanned using modified contact probe of ASD FieldSpec 3 Vis‐NIR Spectrometer (Kusumo et al ., ). Vis‐NIR spectra were analysed using principal component analysis (PCA), and linear discriminant analysis (LDA) was used to detect soil layers where the wavelength bands due to soil organic matter (SOM) content had significantly changed.…”

Predicting pasture root density from soil spectral reflectance: field measurement

“…Visible‐near infrared reflectance spectroscopy (Vis‐NIR) has been used to quantify root density in glasshouse studies (Kusumo et al ., 2009) and soil carbon (C) and nitrogen (N) content in laboratory (Chang & Laird, 2002; Moron & Cozzolino, 2002) and field studies (e.g. Mouazen et al ., 2007; Kusumo et al ., 2008).…”

“…The potential to predict root density from the Vis‐NIRS spectra of soil was demonstrated by Kusumo et al . (2009) using glasshouse‐grown ryegrass plants.…”

VNIR Soil Spectroscopy for Field Soil Analysis

“…To date, studies that address these possible limitations are limited. Chang et al (2005) showed that small increases in moisture content can considerably change the reflectance baseline and increase the peak intensities at 1400 and 1900 nm, although Kusumo et al (2009) effectively used field-moist soil spectra to predict maize root density, SOC, and soil organic nitrogen. More recently researchers (Minasny et al 2011;Nocita et al 2013) have developed methods to remove the effect of soil moisture from soil spectral datasets taken from soils of varying moisture conditions to improve the prediction models.…”

“…The robustness of the calibration is improved by a ten-fold cross-validation procedure on random segments. The number of components to retain in the model has been chosen empirically, based on performance indicators R 2 (coefficient of determination), RPD (ratio of performance deviation), and RMSE (root mean square error of prediction) (see Kusumo et al 2009). A subset of laboratory values are required to provide the calibration set, and the size of the calibration set determines the performance of the prediction model.…”