Evaluation of soil quality for agricultural production using visible–near-infrared spectroscopy

Soil quality (SQ) assessment has numerous applications for agricultural management. Conventional quantification of SQ is based on laboratory analysis and integrative indices that can be costly and time consuming to obtain. A rapid, quantitative method using soil spectra, following the successful process of soil characterization by visible (VIS)–near infrared (NIR) spectroscopy, can provide a robust approach for soil monitoring. To predict specific soil indicator properties and soil quality indices for the productive function of the soil using VIS–NIR spectroscopy, and to evaluate the suitability of spectral data for assessing and monitoring the impact of arable and grassland management in a temperate maritime climate. The study used 40 sites in Ireland under both arable (n=20) and grassland (n=20) management systems. Specific indicators and soil quality indices (SQIs) identified by Askari and Holden (2014) and Askari (2014) were used as the reference standard for estimation using VIS–NIR spectra. Partial least-squares regression was used to predict the indicators and SQIs. SQI was predicted from both spectrally derived indicator values and directly from the soil spectra, and accuracy was assessed by comparison with laboratory and field derived measurements. The indicators of SQ could be predicted with excellent (soil organic carbon and carbon to nitrogen ratio in grassland soils; total nitrogen, carbon to nitrogen ratio, extractable magnesium and aggregate size distribution in arable soils), good (bulk density of