Different types of C-14-labelled substrates, two soluble (glucose and starch) and two particulate (legume and wheat leaves), were incubated in a Vertisol to test the importance of substrate-soil matrix relationships in the processes of soil organic matter decomposition and the location of microorganisms. Mineralized C (CO2 C-12, CO2 C-14) were measured within 66 d of incubation. Sieving and sedimentation procedures were used to fractionate (Light fractions (Lf) > 250 mu m, Lf 50-250 mu m, Heavy fractions (Hf) > 50 mu m, Hf 2-50 mu m, and Hf 0-2 mu m) the soil. Biomass C (C-12 and C-14) in unfractionated soil and in fractions was assayed after 3, 38 and 66 d. Comparisons with an unamended soil (control) were made. Decay rates of substrate C-14 were highest during the first 3 d of incubation. After 66 d, substrate-derived CO2 C-14 represented 63, 64, 59 and 51%, of input C-14 in soils amended with the glucose, starch, legume and wheat, respectively. Unlike C-14, rates of mineralization of C-12 in amended and unamended soils remained more uniform throughout. Total biomass C in soluble substrate-amended soils was similar to that in the control, despite about 60% of total biomass C being derived from C-14 substrate amendments. By contrast, decomposition of particulate substrates increased total biomass C concentration at day 3. There was little or no turnover of C-14 apparent within the first 3 d, as indicated by high (0.60) growth efficiencies (biomass C-14/[biomass C-14, CO2 C-14]). Fraction weights were constant. Irrespective of treatments, the silt-size fraction (Hf 2-50 Irm) was the most abundant (about 51% of total soil weight). This fraction concentrated 65% of the clay fraction as microaggregates. The fraction (Hf > 50 mu m) approximated sand particles ( > 50 mu m). After 3 d, for soils amended with soluble substrate, most (about 65%) of the recovered biomass C-14 was associated with the silt-size fraction (Hf 2-50 mu m) and accounted for 79 and 63% of the total biomass C of that fraction in the cose- and starch-amended soils, respectively. For soils amended with particulate residues, biomass was bimodally distributed, with peak amounts in the silt-size fraction (Hf 2-50 mu m) and the light fraction > 250 mu m (Lf > 250 mu m). In these latter treatments the substrate-derived biomass C-14 associated with the fraction Lf > 250 mu m corresponded broadly to the enhanced total biomass C of the unfractionated soil, when compared with that of the control. Irrespective of substrate amendments, biomass C-14 located in the light fraction (Lf > 250 mu m) had disappeared by 66 d. This decline accounted for more than 50% of biomass C-14 decline from unfractionated soil in particulate plant residue-amended soils. In contrast, in soils amended with soluble substrates, most of the decline in unfractionated soil originated in the silt-size fraction (Hf 2-50 mu m). The nature of the substrate-amendment ensured different sites of microbial activity and turnover, amended particulate residues offering new sites for micro-organisms and soluble compounds stimulating those micro-organisms located within soil matrix (microaggregates 2-50 mu m). (C) 1997 Published by Elsevier Science Ltd. All rights reserved.
