7. Decomposition and Soil Organic Matter Dynamics
over decadal to century intervals. These measures
are particularly important for assessing changes in
soil C pools as a function ofland use change (e.g.,
Haas 1957) and as a function of changes in potential feedbacks in the global C budgets in response
to climate change (e.g., Schlesinger 1990; IPCC
1996). Combustion techniques are less useful for
understanding the role of SOM in ecosystem C dynamics because combustion techniques do not adequately measure C availability.
The availability of C to heterotrophs differs
markedly from one ecosystem to another and can
be independent of total C stores. For example,
ecosystems with identical soil C stores may have
very different soil C cycles because of marked
differences in the availability of C stores to the
heterotrophic community. Soil C takes a wide variety of organic forms, ranging from simple sugars to humified substances with large molecular
weights and complex aromatic structures. These
latter compounds are highly resistant to microbial
attack, especially when protected by association
with stable soil aggregates and clay minerals, and
may thus have turnover times on the order of
millennia.
Efforts to characterize SOM in ecologically
meaningful ways have provided a variety of fractionation methods designed to separate SOM into
functional pools. Classical fractionation procedures separate SOM into humic acid, fulvic acid,
and humin components as a function of relative
solubility in strong acid and base solutions (Sollins
et al. 1999). The humic acid component is soluble
in base but not acid, the fulvic acid component is
soluble in both base and acid, and the humin component is insoluble in base. However, few of these
fractions have been shown to be ecologically relevant across widely different ecosystems, soil types,
and management or disturbance regimes (Duxbury
et al. 1990; Collins et al. 1999; Sollins et al. 1999).
More useful are two simple chemical fractionations: water-soluble C extracts, which appear related to carbon readily available to microbes, and
C released by acid hydrolysis. Acid hydrolysis appears to provide a reasonable measure of the resistant C pool in many soils; Leavitt et al. (1997)
found for a variety of soils that nonhydrolyzable
C is on average 1400 years older than the total C.
Physical fractionation methods use wet sieving,
density flotation gradients, or chemical dispersal to
105
separate SOM into pools of different size and stability classes. Recent work in the U.S. Great Plains
(Cambardella and Elliott 1993) underscores the
potential importance of soil aggregate formation
for protecting SOM. Earlier work in a wide variety
of soils (e.g., Ladd and Amato 1980; Spycher et
al. 1983; Baldock et al. 1990; Christensen 1992)
has shown that SOM particles of different densities
decompose at different rates. Light fractions are
typically comprised of partially decomposed plant
residues with high decomposition rates (Gregorich
and Janzen 1996). That aggregate formation and
the distribution of SOM density classes are under
largely biological control and subject to disturbances imposed by management practices (Oades
1993; Gregorich and Janzen 1996) suggests ecological relevance for these measures.
The proportional composition for any particular
soil appears to be best quantified via microbial
bioassays in which CO 2 release is followed over
long-term incubations of a year or more. This biological fractionation relies on the in situ microbial
and microarthropod community to define ecologically relevant SOM fractions. Biological fractionation schemes have arisen concurrent with the development of quantitative soil C models over the
past two decades. These models have resulted
from efforts to predict the loss rate and the eventual level of soil C following disturbance, and
most now incorporate the concept of mUltiple soil
C pools with different turnover times (Jenkinson
et al. 1987; Parton et al. 1987; Paustian et al.
1992). Usually these models recognize at least
three SOM pools (Fig. 7.1): an active fraction that
turns over on the order of months to years, comprised of recent residues and microbial biomass
and metabolites; a slow fraction that turns over on
the order of decades, comprised of stabilized decomposition products; and a resistant fraction of
highly stabilized, recalcitrant organic matter that
turns over on the order of centuries to millennia.
The curves describing CO 2 release during longterm incubations can be parameterized into two or
more components related to their proportional
composition, as described later. These components
have intrinsic ecological relevance, and a further
value for their application in soil C models (e.g.,
see Fig. 7.1). The application of these models is
becoming increasingly important for evaluating
historical and future global change scenarios (e.g.,
over decadal to century intervals. These measures
are particularly important for assessing changes in
soil C pools as a function ofland use change (e.g.,
Haas 1957) and as a function of changes in potential feedbacks in the global C budgets in response
to climate change (e.g., Schlesinger 1990; IPCC
1996). Combustion techniques are less useful for
understanding the role of SOM in ecosystem C dynamics because combustion techniques do not adequately measure C availability.
The availability of C to heterotrophs differs
markedly from one ecosystem to another and can
be independent of total C stores. For example,
ecosystems with identical soil C stores may have
very different soil C cycles because of marked
differences in the availability of C stores to the
heterotrophic community. Soil C takes a wide variety of organic forms, ranging from simple sugars to humified substances with large molecular
weights and complex aromatic structures. These
latter compounds are highly resistant to microbial
attack, especially when protected by association
with stable soil aggregates and clay minerals, and
may thus have turnover times on the order of
millennia.
Efforts to characterize SOM in ecologically
meaningful ways have provided a variety of fractionation methods designed to separate SOM into
functional pools. Classical fractionation procedures separate SOM into humic acid, fulvic acid,
and humin components as a function of relative
solubility in strong acid and base solutions (Sollins
et al. 1999). The humic acid component is soluble
in base but not acid, the fulvic acid component is
soluble in both base and acid, and the humin component is insoluble in base. However, few of these
fractions have been shown to be ecologically relevant across widely different ecosystems, soil types,
and management or disturbance regimes (Duxbury
et al. 1990; Collins et al. 1999; Sollins et al. 1999).
More useful are two simple chemical fractionations: water-soluble C extracts, which appear related to carbon readily available to microbes, and
C released by acid hydrolysis. Acid hydrolysis appears to provide a reasonable measure of the resistant C pool in many soils; Leavitt et al. (1997)
found for a variety of soils that nonhydrolyzable
C is on average 1400 years older than the total C.
Physical fractionation methods use wet sieving,
density flotation gradients, or chemical dispersal to
105
separate SOM into pools of different size and stability classes. Recent work in the U.S. Great Plains
(Cambardella and Elliott 1993) underscores the
potential importance of soil aggregate formation
for protecting SOM. Earlier work in a wide variety
of soils (e.g., Ladd and Amato 1980; Spycher et
al. 1983; Baldock et al. 1990; Christensen 1992)
has shown that SOM particles of different densities
decompose at different rates. Light fractions are
typically comprised of partially decomposed plant
residues with high decomposition rates (Gregorich
and Janzen 1996). That aggregate formation and
the distribution of SOM density classes are under
largely biological control and subject to disturbances imposed by management practices (Oades
1993; Gregorich and Janzen 1996) suggests ecological relevance for these measures.
The proportional composition for any particular
soil appears to be best quantified via microbial
bioassays in which CO 2 release is followed over
long-term incubations of a year or more. This biological fractionation relies on the in situ microbial
and microarthropod community to define ecologically relevant SOM fractions. Biological fractionation schemes have arisen concurrent with the development of quantitative soil C models over the
past two decades. These models have resulted
from efforts to predict the loss rate and the eventual level of soil C following disturbance, and
most now incorporate the concept of mUltiple soil
C pools with different turnover times (Jenkinson
et al. 1987; Parton et al. 1987; Paustian et al.
1992). Usually these models recognize at least
three SOM pools (Fig. 7.1): an active fraction that
turns over on the order of months to years, comprised of recent residues and microbial biomass
and metabolites; a slow fraction that turns over on
the order of decades, comprised of stabilized decomposition products; and a resistant fraction of
highly stabilized, recalcitrant organic matter that
turns over on the order of centuries to millennia.
The curves describing CO 2 release during longterm incubations can be parameterized into two or
more components related to their proportional
composition, as described later. These components
have intrinsic ecological relevance, and a further
value for their application in soil C models (e.g.,
see Fig. 7.1). The application of these models is
becoming increasingly important for evaluating
historical and future global change scenarios (e.g.,