Speciation of lipid and humic fractions in soils under pine and eucalyptus forest in northwest Spain and its effect on water repellency

Soil organic matter characteristics in extremely water-repellent soils developed under forests of Pinus pinaster and Eucalyptus globulus in Galicia (Northwestern Spain) were analyzed with special emphasis paid in lipid and humic acid fractions. A total of sixteen soils were studied including Leptic Regosols and Leptic Umbrisols developed on granites and schists, and showing extreme water repellency: up to 6 h using water drop penetration time (WDPT) test, i.e., ranging from strongly to extremely hydrophobic (ethanol percentage test, MED). The experimental design involved the measurement of the water repellency as WDPT after the successive removal of lipid and humic fractions by: (i) direct extraction of the soil free lipid with petroleum ether (40–60 °C), (ii) extraction of the ‘fixed’ lipid from the soil residue after 2m H3PO4 treatment and further recovery from the aqueous phase with petroleum ether and (iii) final extraction of humic substances with alkaline solutions. The results showed significant decrease in the WDPT (from class 6 to classes 4 or 5) after removing free lipid. Nevertheless, removal of ‘fixed’ lipid resulted into the most substantial decrease of the WDPT, which occurred on almost all soils (to classes 0–3). This fact is to large extent associated to the simultaneous removal of hydrophobic particulate fractions (free organic matter) which — even after extraction of free lipid — resulted extremely water-repellent in laboratory conditions (>1 h WDPT). Finally, the extraction of humic and fulvic acids was required for the total disappearance of the water repellency (class 0). Regarding vegetation types, lipid removal (free + ‘fixed’ fractions) was significantly more effective in increasing wettability in the case of forest soil samples under pine than under eucalypt. Concerning geological substrate, water repellency in soils under granites remained comparatively more persistent than in soils under schists. After treatment with 2m H3PO4, all samples from soils on granite–eucalyptus remained slightly water-repellent, whereas all samples from soils developed on schist–pine samples became wettable. Multivariate data treatments (multiple regression models, variable ordination by multidimensional scaling, and discriminant analysis) were useful to identify the soil characteristics most significantly associated to its water repellency. These treatments suggest that water repellency in the soils under study is a complex emergent property, reflected in specific patterns depending on vegetation type and geological substrate and arising from the interaction between different soil organizational levels (mainly free lipid, ‘fixed’ lipid, macroscopic particulate organic matter and the concentration and maturity of humic substances).