The structure of the sulfur submodel (Figure 3-6) is similar to the P submodel. The only major difference is that the S model does not include occluded or strongly sorbed pools. There are two mineral pools: sorbed S (SECNDY(3)), and parent S (PARENT(3)). The mineralized S pools (MINERL(*,3)) are specified by soil layer, and summed for the soil rooting depth in TMINRL(3). The sorbed and parent material pools represent the entire soil profile (SOILDEPTH). This structure is shown in the following table:
| Table: Example of the relationships between mineral S pools for a soil containing 3 layers. | ||||||||||||||||||||||||||||||||
|
The main source of S in most soils is the weathering of primary minerals.
Secondary S is formed as a result of adsorption of S on clay minerals.
Organisms in the soil and plant roots take up S from soil solution (MINERL(*,3)) and start the formation of organic
S compounds. The organic component of the S model operates in the same way as
the organic N and P submodels with the C:S ratio of the structural pool being
fixed (500) while the C:S ratios for the active (20-80), slow (90-200) and
passive (20-200) pools vary as a function of the labile S pool (MINERL(*,3)). C:S ratios for surface microbes
are calculated in the same way as the C:N and C:P ratios. The C:S ratios for
the organic components are specified in the fixed parameters set, in file
fix.100. The organic S flows are calculated in the same manner as
the organic N and P flows while the inorganic S flows are functions of
specified rate parameters (fixed parameters PPARMN(3),
PSECMN(3),
PMNSEC(3)) and the moisture and
temperature functions that are used for organic matter decomposition (DEFAC).
The model allows for S fertilization, addition of organic S material, atmospheric deposition (site parameter SATMOS(*)), S in irrigation water (site parameter SIRRI), and accounts for S losses due to crop removal, grazing, leaching of organic S compounds (STREAM(8)), erosion of SOM, and fire.
The S submodel has not been as well tested as the N and P submodels. Parton et al. (1988), Metherell (1992), and Metherell et al. (1993a) describe interactions of S with C, N, and P. The S model could be set up to simulate K dynamics instead of S dynamics if K is a limiting factor in particular soils.