Prediction of humic acids bioactivity using spectroscopy and multivariate analysis

Humic substances (HS) represent the organic material most widespread in nature and have positive effects on plant physiology influencing nutrient uptake and root architecture. Understanding the structure–activity relationship has been considered an essential step to drive technological approaches to produce HS with enhanced plant growth promotion performance. In this study we used solid state nuclear magnetic resonance (CP-MAS 13C NMR) and infrared spectroscopy with diffuse reflectance (DRIFT), a collection of data from humic acid-like substances (HA) obtained from vermicompost produced with different sources of raw material and maturation stage. The bioactivity was evaluated using the ability of HA to induce lateral root emergence sites as well acidification of the liquid medium by treated-maize root seedlings. The structure–activity relationship was done using multivariate analysis including principal component analysis (PCA), principal component regression (PCR) and Spearman correlation joined with NMR and DRIFT results. The main variables (higher loadings) correlated with HA bioactivity were for NMR positive loadings for lignin (56 ppm, 124 ppm, 148 and 153 ppm) and COOH groups (174 ppm) and negative loadings for carbohydrates (64, 75 and 102 ppm). For DRIFT the main positive loadings were aryl stretching and bending (1560, 1480, 860 and 780 cm− 1) probably from lignins; for 2926 cm− 1 and 2852 cm− 1 due to CH asymmetric and symmetric stretching, respectively, probably from long chain fat acids due to the presence of rocking absorption band of (CH2)n with n > 4 at 720 cm− 1 and negative loadings due to carbohydrate moieties (around 1100 cm− 1, CO stretching) and carboxylic acids (1724 cm− 1, CO stretching and 1220 cm− 1, COH stretching and COH bending). The use of spectroscopy data to attempt to relate structure to bioactivity may be of practical interest in developing plant growth promoters using renewed source of humic materials.

► A model for humic acid bioactivity based on chemical properties is proposed
► Multivariate analysis from NMR and DRIFT data can predict humic acids bioactivity.
► We identified the main functionality on humic acids responsible for root growth.