16. Ecosystem Nutrient Balance and Dynamics
al. 1995; Gundersen 1995; Dise and Wright 1995)
in part due to internal differences among ecosystems that now are only beginning to be explored
with more detailed process-level work on internal
cycles and characteristics. Ecosystems may not be
in steady state, or storage may not be constant from
year to year; in these cases, caution must be taken
in interpreting input-output budgets.
Atmospheric Inputs
One of the central concerns in obtaining an accurate
input-output budget is errors associated with measurements of both the atmospheric inputs and the
hydrologic outputs. Concerns with the different
methods of estimating atmospheric deposition to a
forested or otherwise complex watershed terrain are
briefly outlined here and reviewed thoroughly in
Rustad et al. (1994) and Lovett et al (1997). Further
details and discussion of the available measurement
techniques are given in Chapter 17. Most reports in
the literature have used either wet-only deposition
or else bulk deposition to measure total atmospheric inputs to an ecosystem. With the realization
that dry and occult deposition may provide substantial percentages of the total inputs of specific
elements, various techniques were developed to
measure or model dry inputs, none of which have
been generally accepted. Direct measurement techniques such as deposition to surrogate collectors
have known inherent errors, as natural vegetation
surfaces cannot be replicated or even modeled accurately. Current methodologies that use models
and inferential techniques are strongly debated
(Hicks et al. 1980, 1986, 1991). Inferential techniques include the use of towers, met stations, and
surrogate surfaces and filters (see Chapter 17), and
thus are expensive and require a high level of technical expertise compared with other techniques. Because no direct measurement of total deposition to
a canopy has been found, these models cannot be
independently verified, although as the models are
refined and more catchment level budgets are constructed as secondary estimates of inputs, these
techniques may prove to be the most sound. However, canopy type and structure, topographic exposure, elevation, and slope orientation all affect
deposition inputs by wet, dry, and cloud water,
making it difficult to determine deposition inputs
251
to a complex terrain from a single collector at a field
station (Lovett et al. 1997).
Some authors have used throughfall as a surrogate for deposition, as it is an integrator of dry,
cloud, and wet deposition that actually reaches the
vegetation, and a network of throughfall collectors
over a large and variable landscape is more feasible
than a large array of the more expensive deposition
collectors (e.g., Gundersen 1995). However, although canopy uptake and leaching may be minimal for ions such as sulfate (SO~ -), sodium (N a +)
or chloride (CI-), they are significant processes for
ions such as nitrate (NO;) and ammonium (NHt)
(e.g. Lindberg et al. 1986; Friedland et al. 1991;
Lovett et al. 1996) making throughfall estimates of
total deposition less accurate. These processes may
also vary among sites and vegetation types, depending on total rainfall and evergreen/deciduous
differences, making simple corrections inaccurate.
However, across a range of forest types and background N deposition rates at low elevations, Lovett
and Lindberg (1993) calculated a reasonably constant net canopy N uptake of 16% of inputs (,-2
= 0.89), which has been used as a correction factor
for throughfall estimates in similar sites. When an
array of deposition collectors is not feasible, applying such corrections factors to other estimates of
input may be the most accurate technique, with the
understanding that changes in vegetation type, potential inputs from cloud water, etc., will certainly
make the estimates less accurate.
Although no one method to determine atmospheric deposition to an ecosystem has received
complete acceptance by the scientific community,
studies that have calculated a catchment mass balance may offer an independent test for different
methods of estimating dry deposition of various
ions. For example, Hultberg and Grennfelt (1992)
found good agreement between throughfall estimates of salt dry deposition and dry deposition calculated from a watershed mass balance. Rustad et
al. (1994), however, found much higher estimates
of dry deposition inputs of S04 and CI from a catchment budget than from measured throughfall and
canopy mass balance calculations. They suggested
that the discrepancy was due to an error in their
throughfall calculation, as occasional (and thus not
sampled) superdominant conifers in the mixed
hardwood and conifer forests at their study site in
Maine could be extremely efficient scavengers of
al. 1995; Gundersen 1995; Dise and Wright 1995)
in part due to internal differences among ecosystems that now are only beginning to be explored
with more detailed process-level work on internal
cycles and characteristics. Ecosystems may not be
in steady state, or storage may not be constant from
year to year; in these cases, caution must be taken
in interpreting input-output budgets.
Atmospheric Inputs
One of the central concerns in obtaining an accurate
input-output budget is errors associated with measurements of both the atmospheric inputs and the
hydrologic outputs. Concerns with the different
methods of estimating atmospheric deposition to a
forested or otherwise complex watershed terrain are
briefly outlined here and reviewed thoroughly in
Rustad et al. (1994) and Lovett et al (1997). Further
details and discussion of the available measurement
techniques are given in Chapter 17. Most reports in
the literature have used either wet-only deposition
or else bulk deposition to measure total atmospheric inputs to an ecosystem. With the realization
that dry and occult deposition may provide substantial percentages of the total inputs of specific
elements, various techniques were developed to
measure or model dry inputs, none of which have
been generally accepted. Direct measurement techniques such as deposition to surrogate collectors
have known inherent errors, as natural vegetation
surfaces cannot be replicated or even modeled accurately. Current methodologies that use models
and inferential techniques are strongly debated
(Hicks et al. 1980, 1986, 1991). Inferential techniques include the use of towers, met stations, and
surrogate surfaces and filters (see Chapter 17), and
thus are expensive and require a high level of technical expertise compared with other techniques. Because no direct measurement of total deposition to
a canopy has been found, these models cannot be
independently verified, although as the models are
refined and more catchment level budgets are constructed as secondary estimates of inputs, these
techniques may prove to be the most sound. However, canopy type and structure, topographic exposure, elevation, and slope orientation all affect
deposition inputs by wet, dry, and cloud water,
making it difficult to determine deposition inputs
251
to a complex terrain from a single collector at a field
station (Lovett et al. 1997).
Some authors have used throughfall as a surrogate for deposition, as it is an integrator of dry,
cloud, and wet deposition that actually reaches the
vegetation, and a network of throughfall collectors
over a large and variable landscape is more feasible
than a large array of the more expensive deposition
collectors (e.g., Gundersen 1995). However, although canopy uptake and leaching may be minimal for ions such as sulfate (SO~ -), sodium (N a +)
or chloride (CI-), they are significant processes for
ions such as nitrate (NO;) and ammonium (NHt)
(e.g. Lindberg et al. 1986; Friedland et al. 1991;
Lovett et al. 1996) making throughfall estimates of
total deposition less accurate. These processes may
also vary among sites and vegetation types, depending on total rainfall and evergreen/deciduous
differences, making simple corrections inaccurate.
However, across a range of forest types and background N deposition rates at low elevations, Lovett
and Lindberg (1993) calculated a reasonably constant net canopy N uptake of 16% of inputs (,-2
= 0.89), which has been used as a correction factor
for throughfall estimates in similar sites. When an
array of deposition collectors is not feasible, applying such corrections factors to other estimates of
input may be the most accurate technique, with the
understanding that changes in vegetation type, potential inputs from cloud water, etc., will certainly
make the estimates less accurate.
Although no one method to determine atmospheric deposition to an ecosystem has received
complete acceptance by the scientific community,
studies that have calculated a catchment mass balance may offer an independent test for different
methods of estimating dry deposition of various
ions. For example, Hultberg and Grennfelt (1992)
found good agreement between throughfall estimates of salt dry deposition and dry deposition calculated from a watershed mass balance. Rustad et
al. (1994), however, found much higher estimates
of dry deposition inputs of S04 and CI from a catchment budget than from measured throughfall and
canopy mass balance calculations. They suggested
that the discrepancy was due to an error in their
throughfall calculation, as occasional (and thus not
sampled) superdominant conifers in the mixed
hardwood and conifer forests at their study site in
Maine could be extremely efficient scavengers of