Vibrational spectroscopy to determine chemical compounds related to sustainability in soil profiles.

Vibrational spectroscopy techniques are extremely well suited to be used as portable or handheld. Their simplicity, speed, selectivity, and ability to operate without sample preparation make them ideal to be used outside the lab in more difficult environments. Recent evidence supports using visible-near infrared reflectance spectroscopy for sensing soil quality; advantages include low-cost, nondestructive, rapid analysis that retains high analytical accuracy for numerous soil performance measures. Research has primarily targeted agricultural applications (precision agriculture, performance diagnostics), but implications for assessing ecological systems are equally significant. In this work the potential of reflectance spectroscopy using portable instrumentation to determine chemical compounds was evaluated for studying some Mediterranean soils from Sardinia (Italy). These samples had a wide range of soil characteristics due to variations in land use, vegetation cover and specific climatic conditions.
Forty-five soil samples were collected in three different areas (a; b; c) of Sardinia (Italy):
a) in an area representative of Mediterranean agro-forestry systems, located in the North-Eastern Sardinia at Berchidda (40° 47’ 0” N 09° 10’ 0” E) at 320 m above sea level. The soil of the area developed from a granite and was classified as Typic Dystroxerept. Soil texture in the Ap horizon is sandy loam with average pH of 5.7.
b) in a private farm located in a Nitrate Vulnerable Zone in the dairy cattle district of Arborea, (39°47’ N 8°33’ E, 3 m a.s.l.) on the west-coast of Sardinia. The soil was classified as Psammentic Palexeralfs. In the Ap horizon (0-45 cm), the soil of the experimental field had a sandy texture (97% sand) with pH 6.3;
c) in an experimental farm of the Agriculture Faculty of the University of Sassari, located at Ottava (Sassari) in the North-Western Sardinia at 225 m above sea level (40° 46’ 31’’ N 8° 29’ 22’’ E). The soil was a calcareous clay-loam with pH 7.7.
A microNIR 1700 spectrometer (Diessechem Srl, Italy) was used to collect spectra of soil samples in reflectance mode (900 to 1600 nm, 50 scans; 125 reading points). PCA applied to the whole set of spectra allowed the discrimination among the three different soils. On the basis of organic C and total nitrogen contents. Partial Least Squares (PLS) regression with cross validation was used to calculate models in order to quantitatively determine the organic C and the total N contents, found to vary according to the kind of soils.
Using first derivative as pretreatment and considering 7 LV, the organic C content was better predicted (R2 0.849) than total N content, calibrated by using the second derivative as pretreatment and considering 5 LV (R2 0.743).