Geoscience Reference
In-Depth Information
changed as a result of phytoplankton blooms, whereas HMW (~13-150 kDa) material's
fluorescence remained unchanged during bloom conditions. The fractionation scheme
offered greater size resolution than results shown in
Figure 7.2
, and elucidate an impor-
tant size distinction between size classes and DOM sources. Boehme and Wells (
2006
)
stated that T peak fluorescence dominated the “small, dynamic colloidal fraction,” yet this
fraction noticeably lacked marine or terrestrial humic and fulvic fluorescence signals. By
contrast, and as observed in
Figure 7.2E
for the marine permeate (<1 kDa in size), both
terrestrial and marine humic fluorescence was observed for a marine permeate sample that
was collected from a river-dominated margin (the outflow of Chesapeake Bay) and not
after a phytoplankton bloom.
In colloids, an increase in average molecular size may induce hydrophobic character,
resulting in reduced quenching. This could occur if the hydrophobic character of the colloi-
dal material inhibits the interaction of the quencher and fluorophore during the lifetime
of its excited state (Lakowicz,
2006
). Presumably, environmental effects that increase the
hydrophobic nature of DOM could also reduce quenching effectiveness, limiting it to sur-
face fluorophores. Lead et al. (
2006
) employed split-flow thin cell (SPLITT) fractionation
to separate particles and colloids in lake water, an approach somewhat different than ultra-
filtration and FFF, and found the fractions greater than 1 kDa had higher protein fluores-
cence than did humic fractions. They attribute this effect to fluorescence measurement of
tryptophan (or tyrosine) residues on large proteins (or bound to humics or particles) rather
than measurement of free protein. Moreover, they demonstrated that fulvic acid fluores-
cence (C peak) efficiency (normalized to absorption) was lower for their >1 nm (1 kDa)
fraction than for their <1 nm fraction. This would imply some increased quenching of the
C peak fluorescence in the higher molecular size class.
Similar results were obtained for FlFFF separations of DOM from riverine inputs to
the coastal region of the Gulf of Mexico (Stolpe et al.,
2010
) and via ultrafiltration (ca.
12 kDa MWCO) before FlFFF separation for DOM from the humic River Thurso estu-
ary (Batchelli et al.,
2009
). In fact, greater than 50% of the T peak fluorescence in the
Atchafalaya River and Mississippi Bight occurred from DOM fractions greater than 20 nm
in size (ca. 20 kDa: Stolpe et al.,
2010
). In the River Thurso estuary, the authors noted the
appearance of T peak fluorescence in the retentate of their ultrafiltered DOM, but not in
the permeate (Batchelli et al.,
2009
). These riverine and estuarine results are contradictory
to the results of Boehme and Wells (
2006
), but the large amount of T peak fluorescence
found in the higher molecular weight (and size) fractions in the freshwater studies could
be attributable to effects of salinity on DOM fluorescence in estuarine systems (see below;
Boyd and Osburn,
2004
; Batchelli et al.,
2009
).
Boehme and Wells (
2006
) worked on DOM from a largely phytoplankton-dominated
system, in which case terrestrial humic substances are much lower in abundance and con-
sequently the protein fluorescence was most concentrated in the 1-5 kDa fractions. In con-
trast, systems that had substantial humic river influences (or were exclusively freshwater)
exhibited protein fluorescence in the highest size fractions. While free amino acids such
as Trp are smaller molecules, the dominant T peak fluorescence probably occurs from Trp
