CHAPTER 9 . Organic Complexation of Metals in Sea Water
193
oceans (Martin et al. 1991). This is very unexpected because iron is one of the more
abundant elements on earth.
The iron limitation argument, and the fertilization experiments, have thus far been
based on the premise that the solubility of iron in sea water is extremely small
(e.g. Martin and Gordon 1988) and restricted to low nanomolar to subnanomolar levels, the dissolved iron occurring predominantly as hydroxy complexes (Byrne and
Kester 1976; Turner et al. 1981). Laboratory experiments testing the availability of iron
to phytoplanktonic organisms are usually based on excess iron levels, where the available iron is limited by chelation with a synthetic ligand (EDTA) (Brand 1991), and it is
assumed that only the inorganic iron is available to the organisms, the uptake rate being
controlled by complex formation at the transport site into the cells (Hudson and Morei 1990). Although these experiments clearly indicate that iron complexed by EDTA
is not available to microorganisms, there is no evidence that iron complexed by other
organic ligands is also unavailable. Indeed, siderophores are known to increase the
availability of iron to certain marine bacteria (Reid and Butler 1991; Reid et al. 1993).
It is therefore not impossible that organic complexing ligands are released by marine
microorganisms for strategic reasons to convert the iron to a species which is specifically available to them and possibly not to others.
For this reason it is interesting that measurements of the chemical speciation of
iron have shown that this element too is strongly complexed by organic matter in sea
water (Gledhill and van den Berg 1994; Rue and Bruland 1995; van den Berg 1995).
Examples for the distribution of organic ligands for iron are shown for the Mediterranean (Fig. 9.2a) and the Pacific (Fig. 9.2b). There are large differences between these
areas: the Mediterranean receives large amounts of iron as atmospheric inputs which
can vary significantly due to episodic variability. The iron concentration in the water
column of the Mediterranean can therefore be relatively high (2-3 nM) compared to
the subnanomolar levels more typical for open oceanic conditions (Johnson et al.1997).
The ligand concentration was found to be greater than the dissolved iron levels in the
Mediterranean, and was found to show increases in the upper water column associated with, or just below, the chlorophyll maximum. The location of the maximum in
the ligand concentration suggests that it could be the result of breakdown of algal
matter, or production by bacteria, or production by algae, or all of these.
In the Pacific the iron concentrations are much lower, and the ligand concentrations are lower too (Fig. 9.2b): the data were fitted to a model of two complexing
ligands, a strong one with concentrations of 0.5 to 0.6 nM detectable in the upper water
column only due to the higher iron concentrations in the deeper waters, and a weaker
one with concentrations of 1.5 nM in the upper water column to 2.4 nM in the deeper
waters (Rue and Bruland 1995). It is possible that the strong ligand was present throughout the oceanic water column as it could not be detected due to its saturation by the
higher iron concentrations. It is therefore likely that the 2.4 nM of the ligands in the
deeper waters included this strong ligand, and that the weaker ligand occurred at levels between 1.5 and 1.8 nM throughout the water column.
The stability of the iron complexes was very high, with values for the stability constant (K'FeJ+L') of the strong complexes in the Pacific of around 10 23 , whereas for the
ligands in the Mediterranean the constant was just below 10 22 , or about an order of magnitude lower. The organic complexation causes the concentration of inorganic iron to
be lowered by a factor of -500 in the upper water column of the Pacific and through-
193
oceans (Martin et al. 1991). This is very unexpected because iron is one of the more
abundant elements on earth.
The iron limitation argument, and the fertilization experiments, have thus far been
based on the premise that the solubility of iron in sea water is extremely small
(e.g. Martin and Gordon 1988) and restricted to low nanomolar to subnanomolar levels, the dissolved iron occurring predominantly as hydroxy complexes (Byrne and
Kester 1976; Turner et al. 1981). Laboratory experiments testing the availability of iron
to phytoplanktonic organisms are usually based on excess iron levels, where the available iron is limited by chelation with a synthetic ligand (EDTA) (Brand 1991), and it is
assumed that only the inorganic iron is available to the organisms, the uptake rate being
controlled by complex formation at the transport site into the cells (Hudson and Morei 1990). Although these experiments clearly indicate that iron complexed by EDTA
is not available to microorganisms, there is no evidence that iron complexed by other
organic ligands is also unavailable. Indeed, siderophores are known to increase the
availability of iron to certain marine bacteria (Reid and Butler 1991; Reid et al. 1993).
It is therefore not impossible that organic complexing ligands are released by marine
microorganisms for strategic reasons to convert the iron to a species which is specifically available to them and possibly not to others.
For this reason it is interesting that measurements of the chemical speciation of
iron have shown that this element too is strongly complexed by organic matter in sea
water (Gledhill and van den Berg 1994; Rue and Bruland 1995; van den Berg 1995).
Examples for the distribution of organic ligands for iron are shown for the Mediterranean (Fig. 9.2a) and the Pacific (Fig. 9.2b). There are large differences between these
areas: the Mediterranean receives large amounts of iron as atmospheric inputs which
can vary significantly due to episodic variability. The iron concentration in the water
column of the Mediterranean can therefore be relatively high (2-3 nM) compared to
the subnanomolar levels more typical for open oceanic conditions (Johnson et al.1997).
The ligand concentration was found to be greater than the dissolved iron levels in the
Mediterranean, and was found to show increases in the upper water column associated with, or just below, the chlorophyll maximum. The location of the maximum in
the ligand concentration suggests that it could be the result of breakdown of algal
matter, or production by bacteria, or production by algae, or all of these.
In the Pacific the iron concentrations are much lower, and the ligand concentrations are lower too (Fig. 9.2b): the data were fitted to a model of two complexing
ligands, a strong one with concentrations of 0.5 to 0.6 nM detectable in the upper water
column only due to the higher iron concentrations in the deeper waters, and a weaker
one with concentrations of 1.5 nM in the upper water column to 2.4 nM in the deeper
waters (Rue and Bruland 1995). It is possible that the strong ligand was present throughout the oceanic water column as it could not be detected due to its saturation by the
higher iron concentrations. It is therefore likely that the 2.4 nM of the ligands in the
deeper waters included this strong ligand, and that the weaker ligand occurred at levels between 1.5 and 1.8 nM throughout the water column.
The stability of the iron complexes was very high, with values for the stability constant (K'FeJ+L') of the strong complexes in the Pacific of around 10 23 , whereas for the
ligands in the Mediterranean the constant was just below 10 22 , or about an order of magnitude lower. The organic complexation causes the concentration of inorganic iron to
be lowered by a factor of -500 in the upper water column of the Pacific and through-