Stabilisation of soil organic matter by interactions with minerals as revealed by mineral dissolution and oxidative degradation Karin Eusterhues a, *, Cornelia Rumpel a,1 , Markus Kleber b , Ingrid Ko¨gel-Knabner a a Lehrstuhl fu ¨r Bodenkunde, Wissenschaftszentrum Weihenstephan fu ¨r Erna ¨hrung, Landnutzung und Umwelt, Technische Universita ¨t Mu ¨nchen, D-85350 Freising-Weihenstephan, Germany b Institut fu ¨r Bodenkunde und Pflanzenerna ¨hrung, Martin-Luther-Universita ¨t Halle-Wittenberg, D-06099 Halle (Saale), Germany Received 6 January 2003; accepted 28 August 2003 (returned to author for revision 3 April 2003) Abstract Soil organic matter is known to contain a stable fraction with an old radiocarbon age. Size and stabilisation pro- cesses leading to the formation of this old soil carbon pool are still unclear. Our study aims to differentiate old organic matter from young and labile carbon compounds in two acid forest soils (dystric cambisol, haplic podzol). To identify such fractions soil samples were exposed to oxidation with Na 2 S 2 O 8 and to dissolution by hydrofluoric acid (HF). A negative correlation between 14 C activity and carbon release after dissolution of the mineral matrix by HF indicates a strong association of stabilised carbon compounds with the mineral phase. A negative correlation between the 14 C activity and the relative proportion of carbon resistant to oxidation by Na 2 S 2 O 8 shows that young carbon is removed preferentially by this treatment. The fraction remaining after oxidation represents a certain stabilised, long residence time carbon pool. This old fraction comprises between 1 and 30% of the total soil organic carbon in the surface hor- izons, but reaches up to 80% in the sub-surface horizons. Old OC is mainly stabilised by organo-mineral associations with clay minerals and/or iron oxides, whereas intercalation in clay minerals was not found to be important. # 2003 Elsevier Ltd. All rights reserved. 1. Introduction The global carbon cycle is strongly dependent on the turnover of soil organic matter (SOM) (Schlesinger, 1991; Lal, 2001). Part of the organic carbon in soils is easilymineralised,whereasanothercarbonpoolisknown to degrade slowly at timescales from hundreds to thou- sands of years (Oades, 1995). Three key processes are proposed to explain the formation of this passive or long-residence-time SOM fraction: (i) chemical recalci- trance, i.e. stabilisation due to the structural properties of the organic matter, (ii) inclusion of organic matter into aggregates or micropores, leading to physical pro- tection of organic matter from microbial attack and (iii) interaction of carbon compounds with soil minerals (Sollins et al., 1996; Baldock and Skjemstad, 2000; Anderson and Paul, 1984). The contribution and rela- tive importance of these stabilisation mechanisms in different soils are not fully understood. Physical as well as chemical techniques have been used to recover old carbon components (Skjemstad et al., 1996; Falloon et al., 1998; Falloon and Smith, 2000; Christensen, 2001). The chemical approaches include, 0146-6380/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.orggeochem.2003.08.007 Organic Geochemistry 34 (2003) 1591–1600 www.elsevier.com/locate/orggeochem * Corresponding author. Tel.: +49-8161-714195; fax: +49- 8161-714466. E-mail address: eusterhues@wzw.tum.de (K. Eusterhues). 1 Present address: CNRS, Laboratoire de Bioge´ochimie des Milieux Continentaux, Centre INRA Versailles-Grignon, F-78820 Thiverval-Grignon, France.