7. Decomposition and Soil Organic Matter Dynamics
of both the plant residues and soil aggregates. Soils
are typically comprised of a wide variety of particles that differ in size, shape, mineral composition,
and density. Typically, these particles exhibit associated differences in C content and the availability of this C to microbial oxidation. Ecosystemlevel differences in these properties can provide
insight into differences among ecosystems and disturbance regimes with respect to SOM dynamics.
Density fractionation is useful where lighter particles exhibit different C dynamics than heavier particles; most obvious are density differences between organic materials, most of which are light,
and mineral-associated materials, most of which are
heavy (Gregorich and Janzen, 1996). At intermediate densities are a wide range of materials, including amorphous, mineral material adsorbed onto
organic fractions, secondary minerals such as clays
coated with different thicknesses of organic compounds, and soil aggregates of differential stabilities (Turchenek and Oades 1978; Spycher et al.
1983; Tisdall 1996; SoUins et al' 1999). The type
of mineral present can greatly affect decomposition
rates, especially when clays with mixed exchange
characteristics are involved (SoUins et al. 1999).
Soils are most easily separated into density fractions using flotation. The simplest method is to mix
the soil with an inert heavy liquid such as sodium
iodide (NaI) or sodium poly tungstate (NaPT) and
remove the floating fraction (Strickland and SoUins
1987). By sequentially adjusting the density of the
NaI or NaPf, soil fractions of differing densities
can be collected sequentially, usually across an initial range of 1.0 to 2.0 g em - 3. Initial suspensions
can be mixed or sonicated to disperse aggregates of
different stabilities. Each fraction can then be analyzed for C and N content. SoUins et al' (1999)
provide a specific procedure.
Size fractionation (as opposed to density fractionation) is based on the fact that different particle
sizes are differentially reactive because of differences in specific surface areas. For example, claysize particles «2 llm in diameter) have a surface
area that is 10 4 larger than the surface area of sandsize particles «2 mm in diameter): approximately
10 2 versus 10 6 cm 2 /g soil-I, respectively. One
might thus expect the smaller particles to have adsorbed a greater fraction of organic material that
will also be more accessible (because of its larger
surface area) to microbial oxidation.
111
A number of techniques are available for sorting
soil into size classes for subsequent chemical analysis (Gee and Bauder 1986; EUiott et al. 1999).
Efforts to correlate SOM size class fractions with
ecosystem attributes have successfully used withinaggregate particulate organic matter (Cambardella
and EUiott, 1992, 1993) and light fraction (Gregorich and Janzen 1996) components. These coarse
particulate organic matter fractions appear to correlate well with land use management and vegetation changes and with differences in CO2 evolution
in extended laboratory incubations. Carter and
Stewart (1996) review the relationship between soil
structure, various soil C fractions, and SOM storage
and turnover.
Biological Soil Organic Matter Fractions
Biological fractions are defined by the relative rates
of SOM turnover in short- versus long-term soil
incubations. Rates of CO2 release during incubations typically follow an exponential decay curve,
that is, CO2 release is much faster earlier in the
incubation than later. As the more labile C is oxidized in the incubation, the slow-turnover material
is unmasked and both fractions can then be compared with the total C pool to derive fast- and slowturnover SOM fractions (Fig. 7.3). The turnover of
the old resistant pool can be determined by acid
hydrolysis (and if necessary confirmed by C dating)
and is expressed as mean residence time. The
amount of C coming from the old pool is small in
laboratory incubations. The mean residence time
therefore has little effect on the determination of
the dynamics of the active and slow pools, although
the size of the old pool enters into the calculations
and should be estimated independently by acid
hydrolysis.
Laboratory -derived decomposition rate constants
can be converted to mean residence times by the
knowledge that in first order reactions at steady
state, mean residence time is 11k. The laboratory
mean residence times can be converted to field
mean residence times through the use of QIO relationships and knowledge of the field mean annual
temperatures. As noted above, the resistant pool is
measured with acid hydrolysis (Martel and Paul
1974, Leavitt et al. 1997). These three pools analytically define the conceptual pools of the same
names that are incorporated in leading SOM mod-
of both the plant residues and soil aggregates. Soils
are typically comprised of a wide variety of particles that differ in size, shape, mineral composition,
and density. Typically, these particles exhibit associated differences in C content and the availability of this C to microbial oxidation. Ecosystemlevel differences in these properties can provide
insight into differences among ecosystems and disturbance regimes with respect to SOM dynamics.
Density fractionation is useful where lighter particles exhibit different C dynamics than heavier particles; most obvious are density differences between organic materials, most of which are light,
and mineral-associated materials, most of which are
heavy (Gregorich and Janzen, 1996). At intermediate densities are a wide range of materials, including amorphous, mineral material adsorbed onto
organic fractions, secondary minerals such as clays
coated with different thicknesses of organic compounds, and soil aggregates of differential stabilities (Turchenek and Oades 1978; Spycher et al.
1983; Tisdall 1996; SoUins et al' 1999). The type
of mineral present can greatly affect decomposition
rates, especially when clays with mixed exchange
characteristics are involved (SoUins et al. 1999).
Soils are most easily separated into density fractions using flotation. The simplest method is to mix
the soil with an inert heavy liquid such as sodium
iodide (NaI) or sodium poly tungstate (NaPT) and
remove the floating fraction (Strickland and SoUins
1987). By sequentially adjusting the density of the
NaI or NaPf, soil fractions of differing densities
can be collected sequentially, usually across an initial range of 1.0 to 2.0 g em - 3. Initial suspensions
can be mixed or sonicated to disperse aggregates of
different stabilities. Each fraction can then be analyzed for C and N content. SoUins et al' (1999)
provide a specific procedure.
Size fractionation (as opposed to density fractionation) is based on the fact that different particle
sizes are differentially reactive because of differences in specific surface areas. For example, claysize particles «2 llm in diameter) have a surface
area that is 10 4 larger than the surface area of sandsize particles «2 mm in diameter): approximately
10 2 versus 10 6 cm 2 /g soil-I, respectively. One
might thus expect the smaller particles to have adsorbed a greater fraction of organic material that
will also be more accessible (because of its larger
surface area) to microbial oxidation.
111
A number of techniques are available for sorting
soil into size classes for subsequent chemical analysis (Gee and Bauder 1986; EUiott et al. 1999).
Efforts to correlate SOM size class fractions with
ecosystem attributes have successfully used withinaggregate particulate organic matter (Cambardella
and EUiott, 1992, 1993) and light fraction (Gregorich and Janzen 1996) components. These coarse
particulate organic matter fractions appear to correlate well with land use management and vegetation changes and with differences in CO2 evolution
in extended laboratory incubations. Carter and
Stewart (1996) review the relationship between soil
structure, various soil C fractions, and SOM storage
and turnover.
Biological Soil Organic Matter Fractions
Biological fractions are defined by the relative rates
of SOM turnover in short- versus long-term soil
incubations. Rates of CO2 release during incubations typically follow an exponential decay curve,
that is, CO2 release is much faster earlier in the
incubation than later. As the more labile C is oxidized in the incubation, the slow-turnover material
is unmasked and both fractions can then be compared with the total C pool to derive fast- and slowturnover SOM fractions (Fig. 7.3). The turnover of
the old resistant pool can be determined by acid
hydrolysis (and if necessary confirmed by C dating)
and is expressed as mean residence time. The
amount of C coming from the old pool is small in
laboratory incubations. The mean residence time
therefore has little effect on the determination of
the dynamics of the active and slow pools, although
the size of the old pool enters into the calculations
and should be estimated independently by acid
hydrolysis.
Laboratory -derived decomposition rate constants
can be converted to mean residence times by the
knowledge that in first order reactions at steady
state, mean residence time is 11k. The laboratory
mean residence times can be converted to field
mean residence times through the use of QIO relationships and knowledge of the field mean annual
temperatures. As noted above, the resistant pool is
measured with acid hydrolysis (Martel and Paul
1974, Leavitt et al. 1997). These three pools analytically define the conceptual pools of the same
names that are incorporated in leading SOM mod-