Osburn, Del Vecchio, and Boyd
264
DOM fluorescence. This consideration is important for interpreting the change to DOM
fluorescence spectra in estuaries and in coastal waters both of large lakes and of the ocean.
Osburn et al. (2009b) conducted this experiment with Mackenzie River DOM, reproduced
in Figure 7.8. They found that substantial modifications to SF spectra occur that ultimately
remove the humic fluorescence (SF 350 and SF380) but that SF280 fluorescence remained
unchanged. To capture the spectral match of photobleached Mackenzie River DOM fluorescence to Arctic Ocean DOM fluorescence, Pearson correlation coefficients were calculated between SF spectra of river samples before and after photobleaching and SF spectra
of Arctic Ocean DOM. In nearly all cases, photobleaching improved the correlation, often
from r = 0.8 to r = 0.9. In fact, their results also indicated that the UV-A and blue region
wavelengths (>360 nm) were most important for modifying Mackenzie River DOM to
more closely match the spectral signature of Arctic Ocean DOM. A similar modeling
approach by Pullin and Cabaniss (1997) determined that Cuyahoga River water DOM that
was photolabile on the order of 3–7 days could obscure the use of SF as a conservative mixing tracer, producing spectra more similar to those of the Detroit River DOM.
Interactive solution chemistry effects (pH, ionic strength) on DOM fluorescence bleaching are obvious, but trends across environmental systems and with respect to DOM sources
are inconclusive. Reche et al. (1999) showed high CDOM photobleaching rates increasing
with alkalinity, yet acidification was shown to facilitate photobleaching in boreal lakes
(Gennings et al., 2001). Although neither Minor et al. (2006) nor Hefner et al. (2006)
found an effect of salinity on highly absorptive estuarine DOM and Suwannee River humic
acid (SRHA), respectively, Osburn et al. (2009a) showed that longwave photobleaching
of CDOM absorption increased with salinity for ultrafiltered DOM and for SRHA. The
Osburn et al. (2009a) results were corroborated by Grebel et al. (2009) for CDOM absorption, but not for DOM fluorescence, so it appears that DOM fluorescence emission might
respond differently to photobleaching than would DOM absorption.
However, in the SRHA experiment from Osburn et al. (2009a), EEM peak ratio data
do suggest an effect of ionic strength (Figure 7.9, Osburn, unpublished results). The data
show the percent change in A/T, A/C, and A/M peak ratios for SRHA mixed separately
into freshwater, estuarine, and marine permeates and exposed to sunlight. Relative to proteinaceous material, humic material appeared more photolabile (A/T decreased) in lower
salinity water than in seawater. The longer wavelength fluorescent humic material appeared
most photoreactive. The A/C ratio increased with salinity to a greater degree than did the
A/M ratio. This would indicate that the removal of the C peak is sensitive to changes in
salinity, though perhaps for metal–humic interactions or efficiency of peroxide generation
(O’Sullivan et al., 2005). Recently, it has been shown that both LMW DOM A and C peak
fluorescence may be preferentially bleached at lower salinities while HMW DOM A and
C peaks appeared to photobleach more at mid to high salinities (Boyd et al., 2010b). A
four-dimensional PARAFAC model, representing excitation, emission, salinity, pre- and
post-photobleaching showed that T-peak fluorescence increased after exposure while the
A, C, and M peaks were reduced. The changes to peak ratios in Figure 7.9 support the findings of Boyd et al. (2010b) and suggest that salinity may increase photobleaching of DOM
264
DOM fluorescence. This consideration is important for interpreting the change to DOM
fluorescence spectra in estuaries and in coastal waters both of large lakes and of the ocean.
Osburn et al. (2009b) conducted this experiment with Mackenzie River DOM, reproduced
in Figure 7.8. They found that substantial modifications to SF spectra occur that ultimately
remove the humic fluorescence (SF 350 and SF380) but that SF280 fluorescence remained
unchanged. To capture the spectral match of photobleached Mackenzie River DOM fluorescence to Arctic Ocean DOM fluorescence, Pearson correlation coefficients were calculated between SF spectra of river samples before and after photobleaching and SF spectra
of Arctic Ocean DOM. In nearly all cases, photobleaching improved the correlation, often
from r = 0.8 to r = 0.9. In fact, their results also indicated that the UV-A and blue region
wavelengths (>360 nm) were most important for modifying Mackenzie River DOM to
more closely match the spectral signature of Arctic Ocean DOM. A similar modeling
approach by Pullin and Cabaniss (1997) determined that Cuyahoga River water DOM that
was photolabile on the order of 3–7 days could obscure the use of SF as a conservative mixing tracer, producing spectra more similar to those of the Detroit River DOM.
Interactive solution chemistry effects (pH, ionic strength) on DOM fluorescence bleaching are obvious, but trends across environmental systems and with respect to DOM sources
are inconclusive. Reche et al. (1999) showed high CDOM photobleaching rates increasing
with alkalinity, yet acidification was shown to facilitate photobleaching in boreal lakes
(Gennings et al., 2001). Although neither Minor et al. (2006) nor Hefner et al. (2006)
found an effect of salinity on highly absorptive estuarine DOM and Suwannee River humic
acid (SRHA), respectively, Osburn et al. (2009a) showed that longwave photobleaching
of CDOM absorption increased with salinity for ultrafiltered DOM and for SRHA. The
Osburn et al. (2009a) results were corroborated by Grebel et al. (2009) for CDOM absorption, but not for DOM fluorescence, so it appears that DOM fluorescence emission might
respond differently to photobleaching than would DOM absorption.
However, in the SRHA experiment from Osburn et al. (2009a), EEM peak ratio data
do suggest an effect of ionic strength (Figure 7.9, Osburn, unpublished results). The data
show the percent change in A/T, A/C, and A/M peak ratios for SRHA mixed separately
into freshwater, estuarine, and marine permeates and exposed to sunlight. Relative to proteinaceous material, humic material appeared more photolabile (A/T decreased) in lower
salinity water than in seawater. The longer wavelength fluorescent humic material appeared
most photoreactive. The A/C ratio increased with salinity to a greater degree than did the
A/M ratio. This would indicate that the removal of the C peak is sensitive to changes in
salinity, though perhaps for metal–humic interactions or efficiency of peroxide generation
(O’Sullivan et al., 2005). Recently, it has been shown that both LMW DOM A and C peak
fluorescence may be preferentially bleached at lower salinities while HMW DOM A and
C peaks appeared to photobleach more at mid to high salinities (Boyd et al., 2010b). A
four-dimensional PARAFAC model, representing excitation, emission, salinity, pre- and
post-photobleaching showed that T-peak fluorescence increased after exposure while the
A, C, and M peaks were reduced. The changes to peak ratios in Figure 7.9 support the findings of Boyd et al. (2010b) and suggest that salinity may increase photobleaching of DOM