Keywords

Introduction

In the last decades, studies about the influence of no-tillage system on the increase of soil C stocks have been conducted mostly within the top 5- to 20-cm soil layer. In general, these studies verified that each soil has a finite capacity to retain C, and this behaviour is related mainly to soil texture and to the reactivity of mineral surfaces (Stewart et al. 2007, 2008, 2009). Based on this concept, the capacity of soil to sequestrate C would be solely related the formation of organo-mineral complexes. Nevertheless, other mechanisms like self-assembly of organic molecular moieties and the accumulation of C in deeper layers have been also proposed to explain the potential of soil as a C sink (Kleber et al. 2007; Kleber and Johnson 2010). Therefore, the scope of this study was to investigate the influence of soil management system on the C retention capacity of silt and clay fractions in two subtropical oxisols cultivated for 30 years under no-tillage (NT) and conventional tillage (CT) and to estimate its maximal value.

Materials and Methods

Samples (three field replicates) were collected from two long-term experiments of a brown oxisol (BO) and of a red oxisol (RO) in south Brazil under no-tillage (NT) and conventional tillage (CT) in the following depths: 0–2.5, 2.5–5, 5–10, 10–20, 40–60 and 80–100 cm. Samples from native forest (NF) in each site were used as reference of the indigenous vegetation. After removal of particulate organic matter (POM) by sieving, the suspension (<53 μm) was sonicated (1,500 J mL−1), and silt and clay fractions were separated according to Stokes’ law. The silt fraction was further submitted to density fractionation with sodium polytungstate solution (ρ = 2.0 g cm−3) for the removal of the light fraction (LF), resulting in a LF-free silt fraction. The C and N contents of the whole soil and of the fractions were determined by dry combustion. The maximal C retention capacity was estimated from the mathematical models fitted to the plot between the C content in each fraction and \( {C_{\mathrm{ soil}}}\left( {{C_{\mathrm{ fraction}}}={a-b \left/ {{{C_{\mathrm{ soil}}}}} \right.}\ \mathrm{ or}\;{C_{\mathrm{ fraction}}}=a+{b^{*}}\left[ {1-{{\mathrm{ e}}^{{\left( {-{c^{*}}{C_{\mathrm{ soil}}}} \right)}}}} \right]} \right). \) The chemical composition of the soil organic matter (SOM) in each fraction was investigated by FTIR spectroscopy, and the data were analysed by principal component analysis (PCA).

Results and Discussion

In the BO soil, the soil under NT showed higher soil C stocks until 5 cm depth in comparison to CT, while in the RO soil, this difference occurred only in the 0–2.5-cm layer. Regardless the management system, about 29–32% of the total soil C stock of the 0–100 cm depth was accumulated in the top 0–20-cm layer, in both oxisols. The C saturation level estimated for the silt fraction under forest exceeded that of the clay fraction in about 50% for the BO and 200% for the RO (Figs. 1 and 2). After removal of the LF, C saturation of the silt fraction in NF decreased but still remained higher than that in the clay fraction in both soils (Table 1). Similar behaviour was shown by the RO soil under CT and NT, while in the agricultural BO soils, a greater C saturation was found in the clay fraction.

Fig. 1
figure 1

Relation between C content in soil (C soil) and C contents (C fraction) in the fractions of clay (Cc), silt (Cs) and silt fraction free of the light fraction (LF-free Cs) of the brown Oxisol under NF, NT and CT

Full size image
Fig. 2
figure 2

Relation between C content in soil (C soil) and C contents (C fraction) in the fractions of clay (Cc), silt (Cs) and silt fraction free of the light fraction (LF-free Cs) of the red oxisol under NF, NT and CT

Full size image
Table 1 Estimated maximal C retention capacity in the fractions of clay (C clay), silt (C silt) and silt fraction free of the light fraction (LF-free Cs) for the brown oxisol and for the red oxisol under native forest, no-tillage and conventional tillage
Full size table

Conclusions

The retention of SOM in oxisols may be explained by two main processes: the sorption of organic compounds directly on the mineral surface sites and a sorption of SOM on previously sorbed organic coating, leading to a self-association micelle-like structure. The second mechanism plays an important role under CT and NT in the RO soil and under NF in both soils. In the BO soil, both mechanisms seem to contribute equally to C sequestration. Our results elucidate the importance of SOM composition in the C sequestration mechanisms.