15. Biogenic Trace Gas Exchanges
organic carbon (VOC) compounds, and sulfur (S)
gas fluxes.
As with the within-site estimates, regional and
global estimates can be improved by using a stratified sampling approach that partitions the region
on the basis of coarse-scale environmental pattern
(e.g., soil type, climatic type, vegetation type, etc.)
(Matson and Vitousek 1990; Livingston et al. 1988;
Matson et al. 1989; Bartlett et al. 1989, 1993; Matson et al. 1990) or management types (Bouwman
1994; Eichner et al. 1990; Matthews et al. 1994).
Temporal variability must also be accounted for.
Seasonal changes in fluxes driven by changes in
radiation, temperature, precipitation, inundation
patterns, and organic matter inputs are highly significant for many gases. In some systems, periods
of high flux are very short-lived but account for a
majority ofthe annual flux (e.g., Matson et al. 1990;
Whalen and Reeburgh 1992; Matson et al. 1998).
Gas Exchange at the Water-Air Interface
The processes that control exchange of gases between freshwater and coastal marine ecosystems
(including wetlands) and the atmosphere are somewhat more complicated than for soils. The pathways are: (1) diffusive and turbulent flow across
the air-water interface, (2) ebullition (bubbling),
and (3) transfers through emergent aquatic plants.
As with soils, gas transport at the air-water interface is controlled in part by the concentration
gradient between water and air; in aquatic systems,
however, turbulence is also an important factor. For
slightly soluble gases, such as DMS, CH4, and CO2,
diffusive flux (occurring by a combination of molecular diffusive and turbulent diffusive processes)
can be calculated as the gas transfer velocity times
the concentration gradient across the aqueous
boundary layer (the <0.5-mm thick boundary layer
below the air-water interface). The gas transfer velocity is a function of the physical processes (turbulence, molecular diffusion of the gas) and the
viscosity of the water (Liss 1983; MacIntyre et al.
1995) and is empirically related to wind speed, heat
fluxes, and wave characteristics (Wanninkhof
1992). The concentration gradient across the
boundary layer is determined by the sources and
sinks of the trace gases and by variation in solubility with temperature and salinity. Concentrations
are also influenced by highly spatially and tem239
porally variable processes, such as transport of gasladen water from the thermocline, deep waters, hypolimnion, and sediments due to wind mixing,
convective motions, or upwelling. For a discussion
of the relationships between gas transfer velocity
and hydrodynamic and other physical variables,
such as wind speed, wind stress, surface waves, and
turbulence parameters, see MacIntyre et al. (1995).
A number of direct and indirect methods are currently in use to determine gas flux across the waterair interface. Direct methods include the use of
chambers and enclosures, and tracer methods. As
in the soil situation, chambers are placed over the
water surface (either moored or floating with the lip
just below the surface), and change in concentration
is measured in the head-space over time (Frankignolle 1988; Sebacher et al. 1983). In wetlands,
coastal marshes, and mud flats, non-steady state
chambers and flow-through chambers that are floating or sealed to the sediments have been widely
used to estimate exchange of reduced sulfur gases,
especially DMS and hydrogen sulfide (H 2 S) (e.g.,
MacTaggart et al. 1987, Goldan et al. 1987, Morrison and Hines 1990, Castro and Galloway 1991),
and of CH4. The advantage of the chamber method
is that gas fluxes are measured directly rather than
being modeled or extrapolated on the basis of thermodynamic relationships with other known gas
fluxes. The disadvantages are that the chamber isolates the water or sediments from such factors as
wind shear, turbulence, and pressure fluctuations
that affect gas transfer velocity. In addition, singlecomponent systems are particularly problematic in
sediments because chamber placement commonly
causes bubble formation.
The deliberate tracer method evades the problems with disruption of physical processes experienced by chambers and enclosures. In it, a gaseous
tracer such as sulfur hexafluoride (SF 6 ) is added to
the water system and the change in concentration
over time is measured. Where necessary, a nonvolatile tracer is added to the water body with the volatile tracer, in order to allow separation of the effects of dispersion versus gas transfer on change in
concentration. Ideal volatile tracers should be nontoxic, nonreactive, and measurable at low concentrations. See MacIntyre et al. (1995) for a more detailed review of deliberate tracer methods.
A variety of indirect methods to estimate flux
depend on estimation of gas transfer velocity using
organic carbon (VOC) compounds, and sulfur (S)
gas fluxes.
As with the within-site estimates, regional and
global estimates can be improved by using a stratified sampling approach that partitions the region
on the basis of coarse-scale environmental pattern
(e.g., soil type, climatic type, vegetation type, etc.)
(Matson and Vitousek 1990; Livingston et al. 1988;
Matson et al. 1989; Bartlett et al. 1989, 1993; Matson et al. 1990) or management types (Bouwman
1994; Eichner et al. 1990; Matthews et al. 1994).
Temporal variability must also be accounted for.
Seasonal changes in fluxes driven by changes in
radiation, temperature, precipitation, inundation
patterns, and organic matter inputs are highly significant for many gases. In some systems, periods
of high flux are very short-lived but account for a
majority ofthe annual flux (e.g., Matson et al. 1990;
Whalen and Reeburgh 1992; Matson et al. 1998).
Gas Exchange at the Water-Air Interface
The processes that control exchange of gases between freshwater and coastal marine ecosystems
(including wetlands) and the atmosphere are somewhat more complicated than for soils. The pathways are: (1) diffusive and turbulent flow across
the air-water interface, (2) ebullition (bubbling),
and (3) transfers through emergent aquatic plants.
As with soils, gas transport at the air-water interface is controlled in part by the concentration
gradient between water and air; in aquatic systems,
however, turbulence is also an important factor. For
slightly soluble gases, such as DMS, CH4, and CO2,
diffusive flux (occurring by a combination of molecular diffusive and turbulent diffusive processes)
can be calculated as the gas transfer velocity times
the concentration gradient across the aqueous
boundary layer (the <0.5-mm thick boundary layer
below the air-water interface). The gas transfer velocity is a function of the physical processes (turbulence, molecular diffusion of the gas) and the
viscosity of the water (Liss 1983; MacIntyre et al.
1995) and is empirically related to wind speed, heat
fluxes, and wave characteristics (Wanninkhof
1992). The concentration gradient across the
boundary layer is determined by the sources and
sinks of the trace gases and by variation in solubility with temperature and salinity. Concentrations
are also influenced by highly spatially and tem239
porally variable processes, such as transport of gasladen water from the thermocline, deep waters, hypolimnion, and sediments due to wind mixing,
convective motions, or upwelling. For a discussion
of the relationships between gas transfer velocity
and hydrodynamic and other physical variables,
such as wind speed, wind stress, surface waves, and
turbulence parameters, see MacIntyre et al. (1995).
A number of direct and indirect methods are currently in use to determine gas flux across the waterair interface. Direct methods include the use of
chambers and enclosures, and tracer methods. As
in the soil situation, chambers are placed over the
water surface (either moored or floating with the lip
just below the surface), and change in concentration
is measured in the head-space over time (Frankignolle 1988; Sebacher et al. 1983). In wetlands,
coastal marshes, and mud flats, non-steady state
chambers and flow-through chambers that are floating or sealed to the sediments have been widely
used to estimate exchange of reduced sulfur gases,
especially DMS and hydrogen sulfide (H 2 S) (e.g.,
MacTaggart et al. 1987, Goldan et al. 1987, Morrison and Hines 1990, Castro and Galloway 1991),
and of CH4. The advantage of the chamber method
is that gas fluxes are measured directly rather than
being modeled or extrapolated on the basis of thermodynamic relationships with other known gas
fluxes. The disadvantages are that the chamber isolates the water or sediments from such factors as
wind shear, turbulence, and pressure fluctuations
that affect gas transfer velocity. In addition, singlecomponent systems are particularly problematic in
sediments because chamber placement commonly
causes bubble formation.
The deliberate tracer method evades the problems with disruption of physical processes experienced by chambers and enclosures. In it, a gaseous
tracer such as sulfur hexafluoride (SF 6 ) is added to
the water system and the change in concentration
over time is measured. Where necessary, a nonvolatile tracer is added to the water body with the volatile tracer, in order to allow separation of the effects of dispersion versus gas transfer on change in
concentration. Ideal volatile tracers should be nontoxic, nonreactive, and measurable at low concentrations. See MacIntyre et al. (1995) for a more detailed review of deliberate tracer methods.
A variety of indirect methods to estimate flux
depend on estimation of gas transfer velocity using