Dissolved organic matter and high precipitation drive in-situ transition from silandic to aluandic properties

Andosols are commonly subdivided according to silandic and aluandic features. Silandic Andosols are characterised by organic matter (OM) strongly bound to short-range ordered aluminosilicates (SROAS), while aluandic Andosols mainly consist of aluminium-OM complexes (Al-OM complexes). Two theories exist concerning their pedogenesis. One theory argues, that silandic and aluandic properties are direct results of the weathering, assuming two separate lines of genesis. The other theory argues that silandic horizons transform into aluandic over time as parts of a continuous soil forming process. The latter could be caused by dissolved organic matter (DOM) entering the silandic subsoil with the percolating soil solution and promoting the dissolution of SROAS phases by complexing Al. Increasing the loading of DOM with Al will finally result in the formation of insoluble Al-OM complexes. To test the hypothesis of in-situ transition from silandic to aluandic properties in a controlled experiment, we conducted a 20-month percolation experiment with soil material of an Ecuadorian Andosol formed in a homogeneous tephra deposit and now featuring aluandic properties in the top- and silandic properties in the subsoil. Six columns were packed with aluandic material on top of silandic material, water saturated and percolated with litter DOM-solution continuously (percolation rate 8 mm ⋅ h −1 , except for a 9-week flow stop at the beginning of the experiment). In addition, three columns were packed only with aluandic material to gain additional information on the solution entering the silandic material. Among others, silicon (Si) and Al, pH, and dissolved organic carbon (DOC) in the feed and eluate solutions were monitored over a period of 20 months. We modelled the percolation experiment with the convection–dispersion equation as implemented in HYDRUS-1D to estimate the amount of retained DOC in the silandic material. Changes in OC concentration and mineral phases were tracked by analysing the column materials after 0, 8, and 20 months for OC concentrations, oxalate-extractable Al, Si, and iron (Fe) concentrations, and by X-ray diffraction. Our results show that percolation had little to no effect on the aluandic material. However, for the silandic eluate the molar Al:Si ratio was well below the oxalate-extractable Alox:Siox molar ratio of the silandic material itself. This hints at desilification, while Al and OC are retained relative to Si and hence supporting the hypothesis of SROAS dissolution and neo-formation of Al-OM complexes. The latter explained up to 70 % of the massive OC accumulation of 14 mg ⋅ g −1 in the silandic material, while vertical Al-OM transport and sorption played a minor role. This was supported by the HYDRUS-1D modelling, suggesting that sorption of DOM to the silandic material only dominates in the beginning of the percolation, while a DOC sink term, likely representing the formation of Al-OM complexes, operates during later stages of percolation. Overall, our results provide first and compelling experimental evidence of a pedogenic transition from silandic to aluandic Andosols under humid conditions facilitating high percolation through the soil. • Silandic and aluandic soil materials were leached with organic solute solution. • Percolation caused desilification, with Al and OC being retained in the silandic material. • OC retention was driven by sorption only at the beginning of experiment. • OC retention at later stages was due to co-precipitation with Al. • First evidence of a direct transform of silandic into aluandic material.