Iron, being, by weight, the most important of the trace components in algal cells, has long been implicated in the ecology of phytoplank-ton. The two most energy-demanding processes in the cell - photosynthetic carbon reduction and nitrogen reduction - involve the participation of iron-containing compounds deployed in electron transport (such as ferredoxin, nitrogenase) and pigment biosynthesis. Among the recognised direct symptoms of iron deficiency are reduced levels of cytochrome f (Glover, 1977) and the blockage of chlorophyll synthesis (and, where relevant, phycobilins: Spiller et al., 1982). Shortage of iron also impairs the structural assembly of thylakoid membranes (Guikema and Sherman, 1984). Thus, iron-deficient cells are able to harvest relatively fewer photons than iron-replete ones and photon energy is utilised less efficiently. Iron limitation results in poor photosynthetic yields of fixed carbon, lower reductive power and, hence, an impaired growth potential. Iron deficiencies also restrict directly the synthesis of nitrite reductase. For active dinitrogen fixers, the requirement for iron is relatively greater. The synthesis of nitrogenase and the electron power demand for the reduction of nitrogen draws upon the availability of up to 10 times more iron than is needed by cells supplied with assimilable DIN to sustain the equivalent yield of cell carbon (Rueter and Peterson, 1987).
The iron content of phytoplankton cells is reckoned to be between 0.03% and 0.1% of ash-free dry mass, or about 0.1-0.4 mmol Fe : mol C. The problem that cells have in meeting even this modest requirement lies principally in the low solubility of the hydrous ferric oxides that precipitate in aerated, neutral waters. Thus, despite a relative abundance of total iron in fresh waters generally (10-7-10-5 M Fe), most of it is present in flocs and particulates, extremely little (<10-15 M) being in true solution (other than at very low pH: Sigg and Xue, 1994). In the open oceans, the concentration of total iron is generally weaker (10-8-10-7 M Fe). Particular interest has been directed towards the eastern equatorial Pacific, the sub-Arctic Pacific and the Southern Ocean (north of the Antarctic front) where concentrations are considerably lower still (perhaps <10-11 M) and where iron limitation of photosynthesis and growth is demonstrable (Martin, 1992; Martin et al., 1994; and see below).
It has not always been clear just how phy-toplankton cells satisfy their iron needs and at what point these become compromised by an inadequacy of availability. Supplying inorganic iron to cultures and maintaining it in solution requires the inclusion of chelating ligands, such as citrate (Gerloff et al., 1952). It was supposed that this role was fulfilled naturally by the humic or fulvic acids, and that it was suitably imitated by including in media formulations aqueous extracts of soil or by substitution of reproducible solutions of nitrilotriacetic acid (NTA) or trishydroxymethyl-aminomethane (tris). Procedures soon standardised on the use of ethylene diamine tetra-acetic acid (EDTA, usually as its sodium salt) in media in which cultures could be maintained over many generations. EDTA has proved very successful in this context.
It seems that the large molecules of EDTA are too large to be absorbed directly by algae. The role of the chelate is to maintain the source of iron in the medium, stably and accessibly, and which algae are then able to exploit directly, through a process of ligand exchange. The importance of providing an iron source was an incidental confirmation of the procedures for the bioassaying of natural-water samples that became popular in the 1960s and 1970s. The essence of the technique is to grow a test alga, under as near-reproducible laboratory conditions as possible, in water sampled from a given lake or sea under investigation, and in samples of the same water selectively enriched ('spiked') with the suspected regulatory nutrients, separately and in various combinations. Thus, the chemical component that most enhanced the yield of test organism relative to that in the unspiked water was deemed to be the capacity-regulating ('limiting') factor in the original sample (Skulberg, 1964; Maloney et al., 1973). The method would readily confirm previous suspicions about P or N deficiencies but frequently, the tests would point to a direct and previously unsuspected limitation by iron. Alternatively, the effects of N or P spikes were substantially enhanced when iron-EDTA and the relevant spike were added to the medium (Lund et al., 1975). This was true even for water samples from particular lakes previously and deliberately enriched with iron (Reynolds and Butterwick, 1979). The explanation for this behaviour lay almost wholly in the method and its requirement that lake water submitted to bioassay be first fine-filtered of all algal inocula and as many bacteria as possible, prior to the introduction of the test organism. Reynolds et al. (1981) used serial filtrations and intermediate analyses of total iron (TFe) to identify where the loss of iron fertility occurred. Even coarse filtration (50 | m) removed up to one-third of the TFe (as floccular material or finer precipitates on the algae) and glass-filter filtration (pore size 0.45 | m) removed over half of the remainder. From initial TFe concentrations close to 10-5 M (560 |g Fe L-1), the passage of ~10-6 M TFe iron in fine, near-colloidal suspension would nevertheless sustain the subsequent growth of test algae, at least to the point of exhaustion of the conventional N or P additions, without any further enhancement of the iron or the EDTA. On the basis of further experiments, reviewed in Reynolds (1997a), similar results were obtained with iron-starved algae reintroduced into artificial media containing >10-8 M TFe, whereas media containing <10-11 M were consistently too dilute to support any growth of similarly-starved algae. Within this three-order span, results were erratic, either showing some or no growth but which could be stimulated by the addition of Fe-EDTA or, on occasions, by EDTA alone. These performances could not be explained satisfactorily. Some of the variability is attributable to the difficulties of manipulating such low concentrations, when even the impurities present interfere with the nominal interpretation. It would appear, however, that concentrations of residual TFe in the range 10-11-10-10 M may well be exploitable by some algae, provided chelators continue to mediate their availability.
In all these cases, the maintenance of iron in solution by organic chelates is properly emphasised. However, equal emphasis is due to the existence of a mechanism for transferring chelated iron in the medium into the cell. It seems most likely that the uptake and assimilation of iron in the cell relies on reduction of the Fe-chelate at or near the cell surface. In turn, this presupposes that a redox enzyme is produced close to the cell membrane and whose action is to cleave iron from the organic chelates adjacent to the cell surface.
The minimum iron requirements of active nitrogen fixers must, fairly obviously, be relatively higher than those of facultative or obligate users of DIN. It has been suggested that dinitrogen-fixing Cyanobacteria need up to 10 times more iron than algae of the same species growing on DIN at the same rate (Rueter and Peterson, 1987). However, Kustka et al. (2003) have explored the complexities of iron-use efficiency in the diazotrophic Trichodesmium species and calculated the fixed-carbon and fixed-nitrogen quotas required to sustain daily growth rates of 0.1 d-1. The iron use efficiency was such that 1 mol Fe sustained the elaboration of between 2900 and 7700 mol C d-1 (0.13-0.34 mmol Fe : mol C incorporated), thus requiring the supply of 27-48 |imol Fe (mol cell C) d-1. To supply the iron demand of a population equivalent to 0.4-4 x 10-6 mol C L-1 (~0.1-1 |g chla L-1) requires an iron source of 1-2 x10-10 M TFe.
It is relevant to point out that many Cyanobac-teria (though not just the dinitrogen fixers) are able to acquire and transport iron through the production of extracellular iron-binding compounds (called siderophores) that comprise part of their own high-affinity iron-transport systems (Simpson and Neilands, 1976). Production is induced under iron stress and repressed by its relative availability. This ability is said to confer some competitive advantage to Cyanobacteria over other algae (Murphy et al., 1976), though this would seem to apply only to nitrogen fixers under conditions of simultaneous nitrogen stress.
Piecing together the (mainly circumstantial) evidence, it seems scarcely likely that iron exerts any serious regulation on the activities of freshwater phytoplankton. Algae are exposed to relatively high concentrations of TFe supplied by terrestrial soils and that sufficient of the iron is normally maintained in dissolved or colloidal complexes with organic carbon (DOFe) for the carrying capacities of the nitrogen, phosphorus and light to be simultaneously iron-replete. In the sea, however, iron is much more dilute. Apart from fluvial inputs, concentrations are, in part, augmented directly by wet deposition of dust, derived from arid terrestrial (aeolian) sources (Karl, 2002), as well as from deep ocean vents. In much of the sea, organic ligands probably complex sufficient of this iron to ensure its availability to phytoplankton (again, as DOFe) but there remain areas of the ocean where there is just too little iron to avoid a deficiency of supply to autotrophs. It had been suggested that production in the oligotrophic ocean, long supposed to be regulated by nitrogen, would be stimulated by iron additions permitting more nitrogen (and thus carbon) to be fixed (Falkowski, 1997). In the relatively high-nutrient but iron-deficient low-chlorophyll areas of the southern Pacific and the circumpolar Southern Ocean, subject to the IRONEX fertilisation (Martin et al., 1994), it was photosynthesis that was first stimulated by iron addition (Kolber et al., 1994). In a later fertilisation experiment in the Southern Ocean (SOIREE; see Bowie et al., 2001), a pre-infusion concentration of 0.4 nM was raised to give a dissolved iron concentration of 2.7 nM. This was very rapidly depleted within the fertilised patch, to under 0.3 nM. Part of this was due to patch dilution but a distinct biological response confirmed that the biomass limitation was exclusively attributable to iron deficiency. A part of the enhanced iron pool (15-40%) supported the production of autotrophic diatoms and flagellates while the balance persisted (for over 40 days after fertilisation) within the tight cyclic linkages involving pelagic bacteria and microzooplanktic grazers.
We may deduce that, at least for these oceanic locations, the natural iron levels are simply too low (~10-1° M) to support any more autotrophic biomass than they do (i.e. iron availability is absolutely yield limiting and it is not just nitrogen fixation that is constrained). Moreover, the structure of typical iron-limited communities ensures that bioavailable iron is retained, as far as possible, in surface waters.
Was this article helpful?