Hydrophobicity of soils affected by fires: An assessment using molecular markers from ultra-high resolution mass spectrometry

Soil water repellency (SWR) isa physical property due to a complex interaction of factors (e.g., fire, soil organic matter, soil texture) that reduces the soil water infiltration capacity. Traditionally, SWR is attributed to the accumulation and redistribution of hydrophobic compounds within soil profile. To obtain further insight into chemical compounds, which could be associated with SWR, a study was done on coarse (1-2 mm) and fine (< 0.05 mm) granulometric frac-tions of burned and unburned sandy soils under two Mediterranean vegetation biomes from Donana National Park (Spain). The water drop penetration time (WDPT) test was used to assess the SWR. The molecular composition of ex-tracted humic substances from the soil organic matter (SOM) was determined by Fourier transform ion cyclotron res-onance mass spectrometry (FT-ICR/MS). Partial least squares (PLS) regressions showed that the SWR can be predicted (P = 0.006) solely based on the abundances of approximately 1200 common compounds determined by FT-ICR/MS. This model confirmed the significant correlation between a specific SOM molecular composition and the SWR. The comparative analysis revealed that the SWR in the burned samples was significantly (P < 0.05) related to the abun-dance of aromatic and condensed compounds, while in the unburned samples there was a significant influence of ar-omatic hydrocarbons and lignin compounds. In the fine fraction, lipid compounds were significantly associated with the SWR. Contrastingly, the coarse fraction did not show any correlation. Alternatively, soils with a high SWR were significantly related to the presence of lipids and lignin. This analysis showed that combining FT-ICR/MS molecular characterizations with statistical treatments is a powerful approach for exploratory analysis suggesting that the structural features associated with SWR in the studied soils are different depending on the types of vegetation or the soil physical fractions with different particle size.

Automated summary

The study found a close relationship between SWR and the molecular composition of soil organic matter, allowing for the prediction of SWR based on FT-ICR/MS analysis. SWR in burned samples was significantly related to the abundance of aromatic and condensed compounds, while in unburned samples, aromatic hydrocarbons and lignin compounds played a significant role.