Widespread Iron Limitation of Phytoplankton in the South Pacific Ocean

See allHide authors and affiliations

Science  05 Feb 1999:
Vol. 283, Issue 5403, pp. 840-843
DOI: 10.1126/science.283.5403.840


Diel fluorescence patterns were discovered in phytoplankton sampled over 7000 kilometers of the South Pacific Ocean that appear indicative of iron-limiting growth conditions. These patterns were rapidly lost after in situ iron enrichment and were not observed during a 15,000-kilometer transect in the Atlantic Ocean where iron concentrations are relatively high. Laboratory studies of marineSynechococcus sp. indicated that the patterns in the South Pacific are a unique manifestation of iron limitation on the fluorescence signature of state transitions. Results suggest that primary productivity is iron limited not only throughout the equatorial Pacific but also over much of the vast South Pacific gyre.

Evaluating the response of biospheric photosynthesis to alterations in global climate requires an understanding of the geographic distribution of rate-limiting factors for primary production (1). Resolving such distributions in the open ocean has remained at the forefront of oceanographic research for nearly a century (2), but progress has suffered from the lack of physiological diagnostics for particular limiting factors that can be surveyed over vast regions. In situ enrichment experiments (3–5) and molecular probes (6) have now provided critical evidence of iron limitation in two high-nutrient, low-chlorophyll (HNLC) regions, but such techniques are expensive, labor intensive, and limited in their spatial and temporal coverage. Here we describe readily observable diel patterns in photosynthetic parameters that appear to be unique to iron-limited phytoplankton. Our results indicate that iron limitation is not only commonplace in the HNLC equatorial Pacific but also widespread throughout the low-nutrient South Pacific gyre, a vast region classically assumed to be nitrogen limited.

We evaluated photosynthetic parameters from variable fluorescence measurements (7, 8) conducted on phytoplankton collected during two studies in the South Pacific Ocean [OliPac transect, 16°S to 1°N 150°W (1900 km); IronExII transect, 11°S 136°W to 2°S 103°W (5000 km)] and one study in the Atlantic Ocean [AMT-1 transect, 48°N 10°W to 49°S 56°W (15,000 km)]. From these measurements, we derived initial (F 0) and maximal (F m) fluorescence rates, photochemical quantum efficiencies [F v/F m = (F m F o)/F m], and functional absorption cross sections of photosystem II (σPSII) (8). During both the OliPac and the IronExII transects, we collected samples from the low-nutrient, low-chlorophyll South Pacific gyre (SPG) and the HNLC equatorial upwelling (EU) region. Flow cytometer (9) and size-fractionated fluorescence measurements identified Prochlorococcus sp. andSynechococcus sp. as prominent components of the prokaryote-dominated phytoplankton assemblages in the SPG and EU (10). Similar assemblages likewise populate central Atlantic gyres (10).

During OliPac, we observed an unexpected diel pattern inF v/F m throughout the upper 50 m of the water column, which was characterized by a rapid decrease at sunset followed by a reciprocal increase at sunrise. This pattern was further resolved during the IronExII transect and found to occur throughout the SPG and EU (Fig. 1). During each diel cycle, F o andF m exhibited large fluctuations (Fig. 1A) that, in part, reflected protective quenching processes commonly observed under saturating irradiances (11).F v/F mconsistently decreased by 35% to 60% at sunset and subsequently recovered at sunrise, giving a pillared appearance to the transect profile (Fig. 1B). Nocturnal decreases inF v/F m were accompanied by similar decreases in σPSII in the SPG and portions of the EU (Fig. 1C) but were not correlated with changes in chlorophyll concentration.

Figure 1

Diel fluorescence patterns in the South Pacific (A to C) and the Atlantic (Dto F). (A and D) Initial (F 0 = lower gray line) and maximal (F m= upper black line) fluorescence; (B and E) photochemical quantum efficiencies [F v/F m = (F mF 0)/F m]; (C and F) functional absorption cross sections of PSII (σPSII). IX2 indicates beginning of the in situ iron enrichment experiment; EM, eastern margin, NAG, North Atlantic gyre; MUP, Mauritanian upwelling plume; SAG, South Atlantic gyre; SC, subtropical convergence. Curves are based on >13,000 and >21,000 measurements in the South Pacific and Atlantic, respectively (7). Solid and open bars indicate night and day.

Evidence that the nocturnal fluorescence changes were a consequence of iron limitation was first provided during the in situ iron-enrichment experiment conducted immediately after the IronExII transect (4,5). The initial enrichment (2 nM iron) resulted in a rapid decrease in F 0 (Fig. 2A), a small decrease inF m (Fig. 2B), and loss of the extensive nocturnal decreases inF v/F m (Fig. 2C) and σPSII (Fig. 2D). These responses occurred before any change in species composition. Outside the enrichment region, the same diel cycles observed throughout the South Pacific persisted (Fig. 2). These results indicate that the nocturnal decreases inF v/F m and σPSII were a consequence of iron limitation within the enrichment area and suggest that similar iron-limiting conditions prevail throughout the SPG and EU.

Figure 2

Effects of in situ iron enrichment (2 nM iron) on diel fluorescence patterns in the South Pacific. (A) Initial fluorescence (F 0), (B) maximal fluorescence (F m), (C) photochemical quantum efficiencies (F v/F m), (D) functional absorption cross sections of PSII (σPSII). Vertical dash-dot line indicates end of iron enrichment. Solid circles, fluorescence data collected inside the iron enrichment area; open circles, fluorescence data collected outside the enrichment area. Fluorescence patterns from the final day of the transect study (Fig. 1) are shown to the left of the vertical dash-dot line (negative time on x axis). Solid and open bars indicate night and day. Methods are described in (5, 7).

Across the Atlantic Ocean, midday decreases inF v/F m from photoinhibition (12) were often pronounced, but the large nocturnal decreases inF v/F m and σPSII observed in the South Pacific were absent (Fig. 1, E and F). On average,F v/F m in the Atlantic gyres—that is, those regions where phytoplankton assemblages are most comparable to those in the South Pacific (10)—decreased by ≈15% shortly after sunset and then gradually recovered during the night (Fig. 3A). As the aeolian flux of iron to the central Atlantic is one to two orders of magnitude greater than in the South Pacific (13), these results are consistent with an iron-dependent mechanism for the diel patterns in the South Pacific and a general lack of iron limitation in dominant phytoplankton of the Atlantic gyres.

Figure 3

Average diel cycles in photochemical quantum efficiencies (F v/F m) (A) and variable fluorescence (F v =F mF o) (B) for the South Pacific (solid circles) and central Atlantic gyres (open circles). Solid bars indicate night, with vertical dash-dot lines indicating sunset (left) and sunrise (right). AverageF v/F m andF v values were calculated by normalizing each diel cycle to one near sunrise. This normalization removes the influence of changes in phytoplankton biomass on absolute fluorescence values. Shaded area indicates ±1 standard deviation.

Despite the difference between nocturnalF v/F m patterns in the Atlantic and South Pacific (Fig. 3A), changes inF v (that is, F m F o) were remarkably similar (Fig. 3B). However, in the Atlantic gyres, the ≈25% decrease inF v at sunset (Fig. 3B) primarily reflected a decrease in F m, whereas in the South Pacific the same decrease in F v was associated with an increase in background fluorescence, leading to the much larger decrease in F v/F m(Fig. 3A). Nevertheless, similarity between dielF v patterns implied a common physiological mechanism. Coincident changes in σPSII in the South Pacific (Fig. 1C) suggested that this underlying mechanism was an effect of iron limitation on a phenomenon referred to as a state transition (14).

In both eukaryotic and prokaryotic photoautotrophs, the light reactions of photosynthesis require two photosystems (PSII and PSI) for light harvesting and charge separation, along with plastoquinone (PQ) molecules and cytochrome b6–f complexes for proton and electron transport. When photosynthetic cells are exposed to light that preferentially excites PSII, the pool of PQ molecules becomes reduced, which triggers conformational changes in light-harvesting complexes that increase PSI excitation and decrease σPSII by lowering energy transfer to PSII (termed a state 2 transition) (14). These changes may entail a migration of light-harvesting complexes from PSII to PSI [as in eukaryotic algae (14), cyanobacteria (15), and Prochlorophytes (16)] or changes in spillover of excitation energy from PSII to PSI (17) or both. Likewise, when exposed to PSI-specific far red light, the PQ pool becomes oxidized and energy transfer to PSII increases (a state 1 transition).

In prokaryotic photoautotrophs, state 2 transitions also occur upon exposure to darkness (18). This feature of prokaryotes results from common PQ and cytochrome b6–f pools being used for both photosynthesis and metabolism (19). Consequently, photosynthate metabolism at night leads to a reduction of the PQ pool and induction of a state 2 transition.

To investigate the influence of iron limitation on fluorescence parameters during a dark-induced state transition, we grew cultures of marine Synechococcus sp. under nutrient-replete, nitrate-limiting, and iron-limiting conditions (20). As observed in the Atlantic Ocean (Fig. 3), F vdecreased by about 25% andF v/F m decreased by about 15% (k = 0.102 s−1) upon exposure to darkness in both nitrate-limited (Fig. 4A) and nutrient-repleteSynechococcus (21). These changes reflected a decrease in F m and were rapidly reversed (k = 0.026 s−1) upon exposure to PSI-specific far-red light, which induced a state 1 transition (Fig. 4A). Thus, the state 2 transition decreasedF v/F m by increasing energy transfer to PSI, which has a very low fluorescence yield (22).

Figure 4

Changes in variable fluorescence during state transitions in nitrate-limited (A) and iron-limited (B) cultures of marine Synechococcus sp. (20). Data to the left and right of the solid vertical line correspond to dark-induced state 2 transitions and far-red light–induced state 1 transitions, respectively. Open squares,F m; diamonds, F 0 (left axis); solid circles,F v/F m during state 2 transition; open circles,F v/F m during state 1 transition (right axis). For comparison, changes inF v/F m observed in the South Pacific during the first 50 min after sunset (Fig. 3A) are indicated as shaded circles in (B). Results for nutrient-replete cells were similar to those for nitrate-limited cells. Data are normalized to account for differences in absolute values between replicate experiments.F v/F m (circles) was normalized to the value at the beginning of the dark period.F 0 and F m were normalized to the value of F m at the beginning of the dark period.

In iron-limited Synechococcus, the dark-induced state 2 transition likewise resulted in an ≈25% decrease inF v, but it was associated with an increase in background fluorescence (Fig. 4B). Consequently, the decrease inF v/F m was more extensive, following a kinetic rate (k = 0.024 s−1) that was nearly identical to that in the South Pacific (23). Iron-limited Synechococcus also exhibited a PSI-to-PSII ratio that was lower than in nutrient-replete or nitrate-limited cells by a factor of 3. Accordingly, recovery ofF v/F m upon exposure to PSI-specific far-red light was an order of magnitude slower (k = 0.002 s−1) (Fig. 4B).

We thus propose the following underlying mechanism for the divergent Pacific and Atlantic fluorescence patterns. In the South Pacific, iron limitation leads to a decrease in cellular constituents with high iron requirements, such as cytochrome b6–f (5 Fe per complex) and PSI (12 Fe per complex) (24). At sunset, the PQ pool becomes reduced through photosynthate metabolism and a state 2 transition is induced, just as in the Atlantic. However, low PSI-to-PSII ratios in the South Pacific prevent complete association of antennae complexes with PSI. Consequently, during variable fluorescence measurements, antennae complexes decoupled from PSII but not associated with PSI emit absorbed excitation energy, which is observed as an apparent increase in background fluorescence and a decrease inF v/F m. At sunrise, PSI turnover assists in reoxidation of the PQ pool, and the associated state 1 transition leads to a full recovery ofF v/F m.

We propose that iron fluxes in the Atlantic Ocean are sufficient to support relatively high PSI-to-PSII ratios. Thus, nocturnal changes inF v/F m are modest because antennae complexes decoupled from PSII during the state 2 transition at sunset are completely associated with PSI, as in nitrate-limited and nutrient-replete Synechococcus. Likewise, iron enrichment in the South Pacific induces synthesis of iron-dependent cellular constituents (5, 24), which results in rapid decreases in background fluorescence and subsequent increases in F v/F m.

The redox-controlled mechanism thus described is based on iron-dependent alterations in photosystem composition. Similar processes may also occur in eukaryotic photoautotrophs when cytochrome b6–f is sufficiently diminished by iron limitation to cause PQ pool reduction during nocturnal chlororespiration (19), thereby enhancing the general utility of the fluorescence diagnostic. Our results indicate state transitions as the underlying mechanism for the diel fluorescence patterns, although processes such as a back-transfer of electrons from a highly reduced PQ pool to PSII may also be involved (19). Irrespective of the underlying mechanism, a physiological diagnostic for a specific growth-limiting factor proffers a powerful tool for ecological studies and may now be available for iron limitation. Our results evidence widespread iron limitation in both the South Pacific gyre and HNLC equatorial Pacific, thereby greatly expanding the spatial distribution of iron-limited ecosystems in the global ocean.

  • * To whom correspondence should be addressed. Email: behren{at}


View Abstract

Stay Connected to Science

Navigate This Article