Although Guggenberger et al. (1998) postulate predominantly abiotic controls on DOM retention in mineral soil, numerous laboratory studies show that a considerable portion of DOM can be decomposed by microbes. These studies ranged in duration from hours (Yano et al., 1998) to days (Servais et al., 1987; Lucena et al., 1990; Block et al., 1992; Boyer and Groffman, 1996) to a few months (Zsolnay and Steindl, 1991; Qualls and Haines, 1992).
According to Yano et al. (1998), between 12 and 44% of DOC released from the forest floor could be decomposed in solutions by the indigenous microbes. DOC decomposition was also related to nitrogen status: microbial degradability of DOC was higher (43-44%) from plots with a higher N amendment than from the control plots (12-15%). Other experiments have estimated that the biodegradable fraction of DOC is between 10 and 40% (Grøn et al., 1993; Boissier and Fontvieille, 1993; Nelson et al., 1994; Boyer and Groffman, 1996). The range may be partly a consequence of different methods of extracting soil solutions (lysimeter, aqueous soil extracts).
These studies highlight that a distinction between a labile and a refractory DOC fraction is ecologically reasonable (Zsolnay and Steindl, 1991). Zsolnay and Steindl (1991) found in aqueous soil extracts that the solution content of refractory organic carbon is relatively constant, whereas the labile part of DOC ranges between 16 and 68%. Zsolnay (1996) concluded that much of the fresh DOC introduced or produced in the soil has a high substrate value for microbes.
The bioavailable portion of DOC (labile fraction) depends on soil depth (declining with depth) and land use (higher in cornfield than in forest soils) (Boyer and Groffman, 1996). Although DOC concentration, microbial biomass and respiration were significantly higher in organic than in conventional crop farming, the portion of labile DOC was lower in the organic than in conventional soil (Lundquist et al., 1999b). The authors gave two explanations for this phenomenon: (i) the organic soil microbial community may deplete labile DOC to a greater degree than the conventional microbial community and (ii) the type of organic inputs to the two soils may influence the biodegradability of DOC.
Only a few studies have been conducted regarding the biodegradability of DON. According to the work of Qualls and Haines (1992) on soil solution sampled from various depths of an oak forest, DON does not decay faster than DOC. This is consistent with the hypothesis that hydrolysis of organic N is linked to mineralization of DOC rather than occurring selectively in response to the biochemical need for N. Zsolnay (1996), on the other hand, reviewed differences in the biodegradability of water-extractable organic nitrogen and carbon of leaves (Ohta et al., 1986) and agricultural soils (Scherer et al., 1992) and concluded that DON was more biodegradable than DOC. It is not possible at present to resolve these contradicting studies.
Isotope studies support the hypothesis that DOM released in the topsoil is subject to further microbial decomposition. Schiff et al. (1990) and Ludwig and Beese (1997) reported decreasing [delta] C ratios of DOM with increasing soil depth. This can be explained by preferential decomposition of compounds with relatively high [delta] C ratios (carbohydrates, amino acids, pectin, hemicellulose) compared with compounds characterized by relatively low [delta] C ratios (lignin, cellulose). DOM, therefore, seems to be fractionated during soil percolation according to its microbial degradability.
In addition to the high variability of the biodegradable portion of DOM, little is known about the microbial stability of DOM once it is adsorbed onto mineral surfaces. A plausible hypothesis is that the stability of adsorbed organic matter increases because interactions between organic matter and mineral phases decrease the ease of degradation of the organic matter (Sollins et al., 1996; Miltner and Zech, 1998). Another possibility is that the accessibility of sorbed organic matter to microbes and enzymes is diminished (Sollins et al., 1996). Neither of these hypothesis has been tested.
It is reasonable to assume that the chemical properties of DOM affect the rate by which it is decomposed by microbes. Hydrophobic acids, which are preferentially adsorbed in mineral soil horizons, are less accessible to microbial decomposition than hydrophilic components (Qualls and Haines, 1992). Therefore, the more labile DOM components should be preferentially decomposed in soil solution. Mousset et al. (1997), however, did not find any correlation between the proportion of hydrophilic DOM components and the biodegradable part of DOM. Nevertheless, correlations have been found between the biodegradable part of DOM and (i) its elemental composition (Sun et al., 1997) as well as (ii) its specific adsorption in UV light (Zoungrana et al., 1998; Gilbert 1988). Boyer and Groffman (1996) found water-extractable humic acids were more bioavailable than water-extractable fulvic acids. These results challenge the generally accepted opinion that humic acids, because they are high molecular weight organics (Tate, 1987), are more refractory than low molecular weight fulvic acids. Piccolo (1998) suggests that humic substances are associated aggregates of small molecules rather than macromolecules. Therefore, a prediction of the bioavailability of DOM on the basis of its molecular weight seems inappropriate. According to Piccolo (1998), the hydrophobicity of organic matter controls its accumulation and rate of decomposition.
In summary, about 10 to 40% of DOM may be easily decomposed by microbes. At present, it is impossible to quantify the relevance of microbial decomposition in DOM retention in mineral soil horizons. Differences between the microbial stability of DOM in solution and that of DOM adsorbed on mineral soil are not quantified. Microbial decomposition of adsorbed DOM is crucial for the renewal of adsorption sites in mineral soil horizons.