7. Decomposition and Soil Organic Matter Dynamics
macro-, meso-, and megafauna. The microheterotrophs are <100 ).lm and include mainly bacteria,
fungi, nematodes, protozoans, and rotifers; the
mesofauna are generally <2 mm in width and include mites, springtails, diplurans, and enchytraeids; and the macro-fauna include earthworms,
isopods, spiders, molluscs and other organisms >2
mm width. Many of these organisms are detritovores, some are carnivores, and all appear to be
important parts of the soil food web (Swift 1979;
Beare et al. 1995; Coleman and Crossley 1996).
Investigations of decomposition with litter bags
have employed different mesh sizes, sometimes to
experimentally exclude different groups in order to
determine their functional significance for decomposition (e.g., Heath et al. 1964). Generally, however, mesh size has been chosen to optimize the
access of all organisms to entrained litter while
minimizing excessive particle loss in situ. Particle
loss to mineral soil is an important part of decomposition, and too small a mesh size will not only
exclude certain organisms but will also inhibit particle loss. Mesh size may also affect litter microclimate (Witkamp and Olson 1963) if it inhibits
air flow and water flux, and can also alter root
ingrowth.
Although litter bag studies are compromised
somewhat by mesh size limitations, alternatives to
litter bags suffer from equal or greater problems.
For example, litter baskets with mesh bottoms and
mostly-open tops (Stevenson and Dindal 1981;
Blair et al. 1991) allow greater access for macroinvertebrates and better represent the soil surface
microclimate, but also allow additionallitterfall inputs. Tethered material (Witkamp and Olson 1963;
Lang 1974) is completely exposed to the biotic and
abiotic soil surface environment but even large
pieces that fragment from the original material can
be difficult to recover for later weighing and chemical analysis. Thus, litter bags with an appropriate
mesh size remain the preferred means for directly
estimating fine litter mass loss (Harmon et al.
1999).
Indirect measures of estimating decomposition
include estimates of the decomposition rate constant k using litter input: forest floor ratios (Olson
1963), and annual differences in the lowest and
highest stores of litter (e.g., Loomis 1975). These
methods are usually less preferable than direct
means because of their underlying assumption of
107
steady state conditions and the requirement for precise measurements of litter stores, especially where
decomposition rates are low. In some environments, most notably annual cropping systems, these
indirect techniques to estimate mass loss can be
quite accurate.
To allow access to macrofauna, litter bag mesh
size must be >2 mm although I-mm nylon mesh
has been used in most recent litter bag studies. Harmon et al. (1999) recommend a 1.5-mm fiberglass
mesh for high-light environments in which nylon
and other organic materials will be subject to ultraviolet (UV) attack. This is probably a reasonable
compromise that allows access for most mesofauna,
that preserves a near-normal litter layer microclimate, and that avoids excessive loss of large litter
fragments. No mesh size will be free of compromising assumptions, however, and results from any
litter bag study must be interpreted accordingly.
Coleman et al. (1999) recommend specific procedures for assessing macroinvertebrate contributions
to decomposition rates, and these should be consulted for ecosystems in which macroinvertebrates
will substantially affect decay rates.
The size of the litter bag must also be appropriate
to the litter under investigation; 20 X 20 cm is
conventional, but bags may need to be larger for
larger litter or for litter from more diverse plant
communities. Depending on the purpose of the
study, the contents of each litter bag should be proportionately representative of the target ecosystem:
single versus mixed-species litter bags are known
to exhibit different decomposition dynamics (Blair
et al. 1990), and in many plant communities it will
be appropriate to include fine woody debris (e.g.,
twigs). Litter should be freshly senesced.
The number of litter bags to deploy at a site will
vary as a function of expected decomposition rate
and the microclimatic heterogeneity of the site. The
minimum number of bags must be sufficient to describe the decay curve with enough confidence that
the decomposition rate constant k can be adequately
estimated. Usually a geometric sampling interval is
appropriate, that is, more bags are retrieved from
the field earlier than later in the study. Often 3 to 6
collections arranged by season are adequate for the
first year of the study and 2 to 4 collections in
subsequent years. Most studies have examined decomposition dynamics over 1- to 3-year periods;
consequently we know less about late-term decom-
macro-, meso-, and megafauna. The microheterotrophs are <100 ).lm and include mainly bacteria,
fungi, nematodes, protozoans, and rotifers; the
mesofauna are generally <2 mm in width and include mites, springtails, diplurans, and enchytraeids; and the macro-fauna include earthworms,
isopods, spiders, molluscs and other organisms >2
mm width. Many of these organisms are detritovores, some are carnivores, and all appear to be
important parts of the soil food web (Swift 1979;
Beare et al. 1995; Coleman and Crossley 1996).
Investigations of decomposition with litter bags
have employed different mesh sizes, sometimes to
experimentally exclude different groups in order to
determine their functional significance for decomposition (e.g., Heath et al. 1964). Generally, however, mesh size has been chosen to optimize the
access of all organisms to entrained litter while
minimizing excessive particle loss in situ. Particle
loss to mineral soil is an important part of decomposition, and too small a mesh size will not only
exclude certain organisms but will also inhibit particle loss. Mesh size may also affect litter microclimate (Witkamp and Olson 1963) if it inhibits
air flow and water flux, and can also alter root
ingrowth.
Although litter bag studies are compromised
somewhat by mesh size limitations, alternatives to
litter bags suffer from equal or greater problems.
For example, litter baskets with mesh bottoms and
mostly-open tops (Stevenson and Dindal 1981;
Blair et al. 1991) allow greater access for macroinvertebrates and better represent the soil surface
microclimate, but also allow additionallitterfall inputs. Tethered material (Witkamp and Olson 1963;
Lang 1974) is completely exposed to the biotic and
abiotic soil surface environment but even large
pieces that fragment from the original material can
be difficult to recover for later weighing and chemical analysis. Thus, litter bags with an appropriate
mesh size remain the preferred means for directly
estimating fine litter mass loss (Harmon et al.
1999).
Indirect measures of estimating decomposition
include estimates of the decomposition rate constant k using litter input: forest floor ratios (Olson
1963), and annual differences in the lowest and
highest stores of litter (e.g., Loomis 1975). These
methods are usually less preferable than direct
means because of their underlying assumption of
107
steady state conditions and the requirement for precise measurements of litter stores, especially where
decomposition rates are low. In some environments, most notably annual cropping systems, these
indirect techniques to estimate mass loss can be
quite accurate.
To allow access to macrofauna, litter bag mesh
size must be >2 mm although I-mm nylon mesh
has been used in most recent litter bag studies. Harmon et al. (1999) recommend a 1.5-mm fiberglass
mesh for high-light environments in which nylon
and other organic materials will be subject to ultraviolet (UV) attack. This is probably a reasonable
compromise that allows access for most mesofauna,
that preserves a near-normal litter layer microclimate, and that avoids excessive loss of large litter
fragments. No mesh size will be free of compromising assumptions, however, and results from any
litter bag study must be interpreted accordingly.
Coleman et al. (1999) recommend specific procedures for assessing macroinvertebrate contributions
to decomposition rates, and these should be consulted for ecosystems in which macroinvertebrates
will substantially affect decay rates.
The size of the litter bag must also be appropriate
to the litter under investigation; 20 X 20 cm is
conventional, but bags may need to be larger for
larger litter or for litter from more diverse plant
communities. Depending on the purpose of the
study, the contents of each litter bag should be proportionately representative of the target ecosystem:
single versus mixed-species litter bags are known
to exhibit different decomposition dynamics (Blair
et al. 1990), and in many plant communities it will
be appropriate to include fine woody debris (e.g.,
twigs). Litter should be freshly senesced.
The number of litter bags to deploy at a site will
vary as a function of expected decomposition rate
and the microclimatic heterogeneity of the site. The
minimum number of bags must be sufficient to describe the decay curve with enough confidence that
the decomposition rate constant k can be adequately
estimated. Usually a geometric sampling interval is
appropriate, that is, more bags are retrieved from
the field earlier than later in the study. Often 3 to 6
collections arranged by season are adequate for the
first year of the study and 2 to 4 collections in
subsequent years. Most studies have examined decomposition dynamics over 1- to 3-year periods;
consequently we know less about late-term decom-