Hydraulic characterization of earthworm macropore surfaces using miniaturized infiltration data

Abstract

Water repellency related to organic matter (OM) of coatings along the earthworm burrows (EB) walls can facilitate preferential flow due to limiting the water exchange with the soil matrix. The objectives of this work were (i) to characterize the hydraulic behavior of EB wall and the surrounding matrix using a miniaturized infiltrometer; (ii) to compare standard infiltration analyses with BEST-WR algorithm; and (iii) to determine the local spatial distribution of water repellency at the EB wall surface in relation to that of the OM composition. Infiltration experiments were carried out on intact sample surfaces, including the EB wall and the surrounding matrix. Soil samples were collected from Bt and C horizons in a characteristic Haplic Luvisol. Hydraulic conductivity (K) at different water pressure heads (h) and water and ethanol sorptivity (Sw and Se, respectively), water repellency index (RI) and water repellency cessation time (WRCT) were determined from infiltration data. The adapted Beerkan Estimation of Soil Transfer parameters (BEST) algorithm for water-repellent soils (BEST-WR) was used for analysing infiltration data. The potential water repellency index (PWRI), relating C-H with C[dbnd]O functional groups in OM, was determined at the same locations using diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy. Lower values of K and Sw in the coatings as compared to the soil matrix were found related to higher PWI values. Water repellency was also reflected in the shape of the cumulative infiltration curves and in the RI-values. The latter were significantly higher for EB as compared to matrix surfaces. The fittings provided by the BEST-WR algorithm characterizing the hydraulic properties of the macropore surfaces were comparable to those obtained with standard method. The tested methods could help to improve the estimation of parameters for macropore-matrix mass exchange in two-domain models.