66
L.O . Hedin and E.n. Hetherington
Corrections for sea-salt influence had a relatively small «32%) effect
on concentrations of divalent base cations (Ca
2+
and Mg
2
+) in North
American temperate rainforests as well as at CPE (Sections A-D in
Figure 4.3). Monovalent base cations (Na+ and K+) showed greater,
yet varying, reductions for the same locations (Na+: 32-81%; K+:
4-52%). In contrast , sea-salt corrections resulted in strong reductions of
both mono- and divalent base cations for South American temperate
rainforests at CP (section F in Figure 4.3). Concentrations of Mg 2 +, Na+,
and K+ declined by >71 % from these corrections , while the already low
levels of Ca2+ declined by approximately 25% (refer again to section F in
Figure 4.3). Sea-salt corrections had varying effects on SO/- concentrations, with reductions ranging from 6% to 64% for North American
temperate rainforests. In contrast, the South American CP watersheds
showed close to 100% reduction in SO/- from the sea-salt aerosol
correction .
Contributions from Weathering
We evaluated the role of weathering as a source of base cations in
coastal North and South American temperate rainforests by examining
correlations between residual (after sea-salt correction) concentrations of
base cations versus dissolved silica (expressed as SiOz). If silicate
weathering were the major source of residual base cations , we would
expect residual cation concentrations to correlate positively with variations
in SiOz.
When data from all coastal watersheds were considered together,
residual concentrations of Ca?", Mg z +, and Na" were, in fact, positively
correlated with concentrations of SiOz (see Figure 4.4). (Ca z +: r
Z = 0.64,
P < 0.001; Mg'": r
Z
= 0.59, P < 0.001; Na+: r = 0.22; P < 0.002; all are
log-log regressions.) The slopes of these relationships were consistently
less than 1.0 (Ca?" = 0.5; Mg z+ = 0.5; and Na+ = 0.4) , indicating that
on a linear scale, the non-sea-salt supply of base cations did not increase
proportionately as a function of SiOz. We do not present results for K+ in
Figure 4.4, since some residual K+ levels were slightly negative and
therefore could not be log-transformed . However , evaluation of non-logtransformed data showed a positive and highly significant (p < 0.001;
r = 0.30) relationship between residual K+ and SiOz. It is particularly
interesting that watersheds with low levels of residual cations were exclusively associated with very low SiOz (ca. 0.9-3.0mgl1 ) (refer to
Figure 4.4). Such levels of SiOz are low even when compared to waters
that drain lateritic terranes of central Amazonia (3-12 mg1-\ Kronberg
& Melfi, 1987).
L.O . Hedin and E.n. Hetherington
Corrections for sea-salt influence had a relatively small «32%) effect
on concentrations of divalent base cations (Ca
2+
and Mg
2
+) in North
American temperate rainforests as well as at CPE (Sections A-D in
Figure 4.3). Monovalent base cations (Na+ and K+) showed greater,
yet varying, reductions for the same locations (Na+: 32-81%; K+:
4-52%). In contrast , sea-salt corrections resulted in strong reductions of
both mono- and divalent base cations for South American temperate
rainforests at CP (section F in Figure 4.3). Concentrations of Mg 2 +, Na+,
and K+ declined by >71 % from these corrections , while the already low
levels of Ca2+ declined by approximately 25% (refer again to section F in
Figure 4.3). Sea-salt corrections had varying effects on SO/- concentrations, with reductions ranging from 6% to 64% for North American
temperate rainforests. In contrast, the South American CP watersheds
showed close to 100% reduction in SO/- from the sea-salt aerosol
correction .
Contributions from Weathering
We evaluated the role of weathering as a source of base cations in
coastal North and South American temperate rainforests by examining
correlations between residual (after sea-salt correction) concentrations of
base cations versus dissolved silica (expressed as SiOz). If silicate
weathering were the major source of residual base cations , we would
expect residual cation concentrations to correlate positively with variations
in SiOz.
When data from all coastal watersheds were considered together,
residual concentrations of Ca?", Mg z +, and Na" were, in fact, positively
correlated with concentrations of SiOz (see Figure 4.4). (Ca z +: r
Z = 0.64,
P < 0.001; Mg'": r
Z
= 0.59, P < 0.001; Na+: r = 0.22; P < 0.002; all are
log-log regressions.) The slopes of these relationships were consistently
less than 1.0 (Ca?" = 0.5; Mg z+ = 0.5; and Na+ = 0.4) , indicating that
on a linear scale, the non-sea-salt supply of base cations did not increase
proportionately as a function of SiOz. We do not present results for K+ in
Figure 4.4, since some residual K+ levels were slightly negative and
therefore could not be log-transformed . However , evaluation of non-logtransformed data showed a positive and highly significant (p < 0.001;
r = 0.30) relationship between residual K+ and SiOz. It is particularly
interesting that watersheds with low levels of residual cations were exclusively associated with very low SiOz (ca. 0.9-3.0mgl1 ) (refer to
Figure 4.4). Such levels of SiOz are low even when compared to waters
that drain lateritic terranes of central Amazonia (3-12 mg1-\ Kronberg
& Melfi, 1987).