17. Deposition of Nutrients and Pollutants to Ecosystems
isotope ratios or concentrations). In contrast, the
more traditional measures of deposition fluxes derive from extensive quantities (e.g., mass fluxes of
nutrients) that depend directly on the size and
boundaries of the system in question. Recent summaries by Kendall and McDonnell (1998), Kendall
et al. (1995), and Macko and Ostrom (1994) discuss
some of the prospects and difficulties of using stable isotopes to infer nutrient and element fluxes
across the atmosphere-land interface.
For example, natural abundances of Sr isotopes
have been used to understand Ca 2 + cycling in ecosystems (Ca 2 + isotopes are rarely used due to analytical difficulties, and due to less predictable patterns in biological and geological systems).
Graustein and Armstrong (1983) took advantage of
local differences in isotopic signals of strontium
(Sr) associated with bedrock (assumed to reflect
weathering inputs) and with atmospheric dust. By
analyzing the natural abundance 87Sr/86Sr ratios in
plant materials, Graustein and Armstrong could
then deduce the exact mixture of the two putative
sources (weathering vs. atmospheric sources)
needed to produce the isotopic ratio measured in
plant tissues. In this manner it is possible, if local
conditions permit a clear separation among putative
end-members, to quantify the degree to which atmospheric Sr deposition (and by extension other
base cations) contribute to internal nutrient cycles
of entire ecosystems. The technique has now been
more widely employed in forested ecosystems
(e.g., Aberg et al. 1989; Miller et al. 1993; Bailey
et al. 1996; Dambrine et al. 1997; Kennedy et al.
1998). However, it should be cautioned that this
approach, like other natural abundance isotope
techniques, is associated with uncertainties and
site-specific challenges. The approach assumes that
the two end-members (weathering vs. atmospheric
sources) are isotopically distinct, homogeneously
mixed, and unvarying in time and space. Unfortunately, these assumptions are likely to be violated
in many ecosystems. For example, the atmospheric
end-member is typically neither homogeneous nor
unvarying, but originates from heterogeneous
mixtures of several sources that differ in 87Sr/86Sr
signatures and that vary in relative contribution
over ecological and geological time scales. Weathering is equally complex, often involving different
combinations of primary and secondary minerals
that may vary in both weathering rate and isotopic
273
signature of weathered Sr. A final difficulty is that
Sr typically does not behave identically to other
cations, and that the magnitude of these differences
among cations may, in tum, vary from site to site.
Nevertheless, the approach can be quite useful in
locations where end-members are clearly different,
and where the underlying assumptions can be carefully examined.
Even greater difficulties exist in using ISN isotopic signatures to study atmospheric N deposition
to ecosystems. First, ISN signatures are quite variable in wet deposition (Durka et al. 1994; Macko
and Ostrom 1994). Second, ISN signatures within
forest ecosystems range greatly due to isotopic
fractionation (systematic change in isotopic ratios)
associated with the processes of N uptake, mineralization, denitrification, and other biological
transformations (e.g., Hedin 1994; Macko and Ostrom 1994). To overcome the difficulties associated with such internal fractionation, Durka et al.
(1994) employed a dual-label eSN and 18 0) isotope approach to trace atmospheric NO; inputs to
forests in Germany. By simultaneously measuring
the ISN and 18 0 signatures in NO; in streams
draining forests, they were able to more clearly
separate the isotopic signatures of atmospherically
deposited NO; vs. NO; produced by nitrification
within the forest ecosystem. This study is an example of how relatively advanced isotopic analyses can be used to understand how atmospheric
deposition impacts ecosystems (see Hedin 1994).
In a different application of watershed-scale
tracers, Hedin et al. (1995) and Hedin and Hetherington (1996) used the dominance of sea-salt
aerosols in the atmospheres of unpolluted temperate forests in southern Chile to quantify the role
of atmospheric cation deposition. By assuming
that chloride (Cl-) acts as a conservative tracer in
forested watersheds, Hedin et al. (1995) used seasalt ratios of elements to CI- to subtract out the
contribution of atmospheric deposition to element
outputs in watershed streams. Such subtraction removed the majority of solutes in streams, indicating that these ecosystems depend closely on atmospheric deposition (also see Chadwick et al.
[1999]).
Summary and Prospects
The past four decades have seen rapid developments in our abilities to measure atmospheric nutrient inputs to ecosystems. While there still is
isotope ratios or concentrations). In contrast, the
more traditional measures of deposition fluxes derive from extensive quantities (e.g., mass fluxes of
nutrients) that depend directly on the size and
boundaries of the system in question. Recent summaries by Kendall and McDonnell (1998), Kendall
et al. (1995), and Macko and Ostrom (1994) discuss
some of the prospects and difficulties of using stable isotopes to infer nutrient and element fluxes
across the atmosphere-land interface.
For example, natural abundances of Sr isotopes
have been used to understand Ca 2 + cycling in ecosystems (Ca 2 + isotopes are rarely used due to analytical difficulties, and due to less predictable patterns in biological and geological systems).
Graustein and Armstrong (1983) took advantage of
local differences in isotopic signals of strontium
(Sr) associated with bedrock (assumed to reflect
weathering inputs) and with atmospheric dust. By
analyzing the natural abundance 87Sr/86Sr ratios in
plant materials, Graustein and Armstrong could
then deduce the exact mixture of the two putative
sources (weathering vs. atmospheric sources)
needed to produce the isotopic ratio measured in
plant tissues. In this manner it is possible, if local
conditions permit a clear separation among putative
end-members, to quantify the degree to which atmospheric Sr deposition (and by extension other
base cations) contribute to internal nutrient cycles
of entire ecosystems. The technique has now been
more widely employed in forested ecosystems
(e.g., Aberg et al. 1989; Miller et al. 1993; Bailey
et al. 1996; Dambrine et al. 1997; Kennedy et al.
1998). However, it should be cautioned that this
approach, like other natural abundance isotope
techniques, is associated with uncertainties and
site-specific challenges. The approach assumes that
the two end-members (weathering vs. atmospheric
sources) are isotopically distinct, homogeneously
mixed, and unvarying in time and space. Unfortunately, these assumptions are likely to be violated
in many ecosystems. For example, the atmospheric
end-member is typically neither homogeneous nor
unvarying, but originates from heterogeneous
mixtures of several sources that differ in 87Sr/86Sr
signatures and that vary in relative contribution
over ecological and geological time scales. Weathering is equally complex, often involving different
combinations of primary and secondary minerals
that may vary in both weathering rate and isotopic
273
signature of weathered Sr. A final difficulty is that
Sr typically does not behave identically to other
cations, and that the magnitude of these differences
among cations may, in tum, vary from site to site.
Nevertheless, the approach can be quite useful in
locations where end-members are clearly different,
and where the underlying assumptions can be carefully examined.
Even greater difficulties exist in using ISN isotopic signatures to study atmospheric N deposition
to ecosystems. First, ISN signatures are quite variable in wet deposition (Durka et al. 1994; Macko
and Ostrom 1994). Second, ISN signatures within
forest ecosystems range greatly due to isotopic
fractionation (systematic change in isotopic ratios)
associated with the processes of N uptake, mineralization, denitrification, and other biological
transformations (e.g., Hedin 1994; Macko and Ostrom 1994). To overcome the difficulties associated with such internal fractionation, Durka et al.
(1994) employed a dual-label eSN and 18 0) isotope approach to trace atmospheric NO; inputs to
forests in Germany. By simultaneously measuring
the ISN and 18 0 signatures in NO; in streams
draining forests, they were able to more clearly
separate the isotopic signatures of atmospherically
deposited NO; vs. NO; produced by nitrification
within the forest ecosystem. This study is an example of how relatively advanced isotopic analyses can be used to understand how atmospheric
deposition impacts ecosystems (see Hedin 1994).
In a different application of watershed-scale
tracers, Hedin et al. (1995) and Hedin and Hetherington (1996) used the dominance of sea-salt
aerosols in the atmospheres of unpolluted temperate forests in southern Chile to quantify the role
of atmospheric cation deposition. By assuming
that chloride (Cl-) acts as a conservative tracer in
forested watersheds, Hedin et al. (1995) used seasalt ratios of elements to CI- to subtract out the
contribution of atmospheric deposition to element
outputs in watershed streams. Such subtraction removed the majority of solutes in streams, indicating that these ecosystems depend closely on atmospheric deposition (also see Chadwick et al.
[1999]).
Summary and Prospects
The past four decades have seen rapid developments in our abilities to measure atmospheric nutrient inputs to ecosystems. While there still is