Abstract. Extreme sensitivity of soil organic carbon (SOC) to climate and land use change warrants further research in different terrestrial ecosystems. The aim of this study was to investigate the link between aggregate and SOC dynamics in a chronosequence of three different land uses of a south Chilean Andisol: a second growth Nothofagus obliqua forest (SGFOR), a grassland (GRASS) and a Pinus radiata plantation (PINUS). Total carbon content of the 0-10 cm soil layer was higher for GRASS (6.7 kg C m −2 ) than for PINUS (4.3 kg C m −2 ), while TC content of SG-FOR (5.8 kg C m −2 ) was not significantly different from either one. High extractable oxalate and pyrophosphate Al concentrations (varying from 20.3-24.4 g kg −1 , and 3.9-11.1 g kg −1 , respectively) were found in all sites. In this study, SOC and aggregate dynamics were studied using size and density fractionation experiments of the SOC, δ 13 C and total carbon analysis of the different SOC fractions, and C mineralization experiments. The results showed that electrostatic sorption between and among amorphous Al components and clay minerals is mainly responsible for the formation of metal-humus-clay complexes and the stabilization of soil aggregates. The process of ligand exchange between SOC and Al would be of minor importance resulting in the absence of aggregate hierarchy in this soil type. Whole soil C mineralization rate constants were highest for SGFOR and PINUS, followed by GRASS (respectively 0.495, 0.266 and 0.196 g CO 2 -C m −2 d −1 for the top soil layer). In contrast, incubation experiments of isolated macro organic matter fractions gave opposite results, showing that the recalcitrance of the SOC decreased in another order: PINUS>SGFOR>GRASS. We deduced that electrostatic sorption processes and physical protection of SOC in soil aggregates were the main processes determining SOC stabilization. As a result, high aggregate carbon concentraCorrespondence to: D. Huygens (dries.huygens@ugent.be) tions, varying from 148 till 48 g kg −1 , were encountered for all land use sites. Al availability and electrostatic charges are dependent on pH, resulting in an important influence of soil pH on aggregate stability. Recalcitrance of the SOC did not appear to largely affect SOC stabilization. Statistical correlations between extractable amorphous Al contents, aggregate stability and C mineralization rate constants were encountered, supporting this hypothesis. Land use changes affected SOC dynamics and aggregate stability by modifying soil pH (and thus electrostatic charges and available Al content), root SOC input and management practices (such as ploughing and accompanying drying of the soil).