Estimating Soil Properties and Nutrients by Visible and Infrared Diffuse Reflectance Spectroscopy to Characterize Vineyards

Summary

This study applied visible, near, and shortwave infrared reflectance spectroscopy as a cost-effective and rapid method to characterise soil properties across four Spanish vineyards. Partial least squares regression models achieved excellent predictive performance for pH, electrical conductivity, and phosphorus (R² > 0.92), and moderate performance for sand, nitrogen, and potassium (R² 0.69–0.77), with shortwave infrared data and standard normal variate plus detrending transformation yielding the strongest models. The findings indicate that VIS-NIR-SWIR spectroscopy holds promise as a practical tool for soil characterisation in precision viticulture applications, though further validation is recommended.

UK applicability

The spectroscopic method demonstrated here could be adapted for UK vineyard and wider horticultural soil assessment, though the calibrations developed in Spanish soils would likely require validation across different UK soil types and climates. The portable equipment and rapid analysis time suggest potential for adoption in UK precision agriculture programmes, provided appropriate soil-specific calibration sets are developed.

Key measures

R² values for PLSR model predictions of soil pH, electrical conductivity, phosphorus, sand, nitrogen, and potassium; spectral reflectance data across VIS (350–700 nm), NIR (701–1000 nm), SWIR (1001–2500 nm), and full range (350–2500 nm); comparison of pistol grip versus contact probe measurement setups

Outcomes reported

The study evaluated the performance of visible, near, and shortwave infrared (VIS-NIR-SWIR) reflectance spectroscopy in predicting soil properties across four vineyards in central and north-western Spain. Partial least squares regression models were developed to predict pH, electrical conductivity, phosphorus, sand, nitrogen, and potassium content from spectral reflectance data.

Topic tags