in oxidation states. Sulfur occurs in soil both in organic and inorganic forms. However, most of the soil S Sulfur K-edge X-ray absorption spectroscopy (XANES) was used (95%) in temperate and tropical ecosystems is present to identify S oxidation states and assess the impact of land use changes on the amount, form, and distribution of organic S in particle-size in organic forms (Janzen and Ellert, 1998; Saggar et al., separates and their humic substance extracts. Soil samples (0-10 cm) 1998; Solomon et al., 2001a). Knowledge of the specia- were collected from natural forest, tea plantations, and cultivated tion of S in organic materials could, therefore, provide fields at Wushwush and from natural forest, Cupressus plantations a clearer understanding of the amount, form, and bio- and cultivated fields at Munesa sites in Ethiopia. Sulfur XANES geochemical transformations of S in the ecosystem. Our spectra measured directly from the size separates could not be quanti- knowledge of the biogeochemical S cycling is, however, tatively analyzed due to high background noise. However, qualitative severely limited by relatively crude analytical tech- comparison of spectra from size separates and their humic extracts niques used in S measurements and speciation. Most of were very similar and thus provides a characteristic fingerprint of S the present studies on soil organic S are based on a in mineral soils. X-ray absorption near-edge spectroscopy showed the fractionation technique using HI reduction of organic presence of most reduced (sulfides, disulfides, thiols, and thiophenes), intermediate (sulfoxides and sulfonates) and highly oxidized S (ester- S compounds in soil (Tabatabai, 1982; Kowalenko, SO4-S) forms. Sulfur in intermediate oxidation states was dominant 1993a,b). The biochemical characterization of organic (39-50%; where 66-96% of it being sulfonate S) in humic extracts S as ester-SO4-S (HI-reducible S) and C-bonded S has from clay, while highly oxidized S dominated (40-56%) the silt spectra. advanced the understanding of the organic S cycle in Concentrations of C-bonded and ester-SO4-S extracted by the HI soil (e.g., McGill and Cole, 1981; Janzen and Ellert, fractionation did not correlate with those from XANES (ester-SO4-S 1998; Lehmann et al., 2001; Solomon et al., 2001a). How- revealed by XANES vs. HI-fractionation, r 0.23; P 0.001). A ever, this technique is only an indirect method involving major shift following land use changes occurred in the most reduced differential reduction of organic S compounds to H 2S and intermediate S species. Their proportion decreased in the order: and is, therefore, not an entirely satisfactory procedure natural forests plantations cultivated fields. In contrast, highly to directly speciate S within complex organic functional oxidized S increased in the order: natural forests plantations cultivated fields at both sites. Our results indicated that C-bonded S groups. Moreover, this chemical reduction method can- (most reduced and intermediate S) may represent the more labile not identify intermediate oxidation states of S in organic forms of organic S compounds compared with ester-SO4-S. Therefore, matter and consequently almost no information is avail- S K-edge XANES has a significant potential to evaluate the influence able about their turnover rates. Other studies have at- of anthropogenic changes on the nature and distribution of S and to tempted to speciate S into its functional groups using follow its dynamics in terrestrial ecosystems.