Early summer iron limitation of phytoplankton photosynthesis in the Scotia Sea as inferred from fast repetition rate fluorometry
Abstract
[1] We describe variability in the phytoplankton physiological status in the west Scotia Sea of the Southern Ocean in the early austral summer prior to and at the beginning of a phytoplankton bloom. This area is characterized by high concentrations of major nutrients (such as nitrate, phosphate, and silicate), but exhibit chronically low concentrations of dissolved iron and deep vertical mixing. Using a fast repetition rate fluorometry, we measured photosynthetic characteristics of phytoplankton in the euphotic zone. These measurements provide an express diagnostic of the effects of environmental factors, including iron limitation, on photosynthetic processes. The quantum yields of photochemistry in Photosystem II (Fv/Fm) in subsurface phytoplankton were, on average, 40% lower than the maximum values for nutrient-replete communities. Higher values of Fv/Fm were observed in the frontal mixing zone that may have been caused by the induction of iron into the euphotic layer. Our results suggest physiological signatures of iron limitation of photosynthesis in the western Scotia Sea in the early austral summer. The data imply that, together with light conditions and grazing pressure, iron availability may be important for phytoplankton growth in the Scotia Sea even in the early summer.
Key Points
- Characterized physiological status of phytoplankton in the western Scotia Sea.
- Revealed physiological signatures of iron limitation in early summer
- Discuss the importance of iron availability and physical forcing
1. Introduction
[2] Despite permanently high concentrations of inorganic macronutrients [Garcia et al., 2006; Mitchell and Holm-Hansen, 1991] and seasonally high light inputs, the majority of surface waters in the Southern Ocean (SO) are characterized by low chlorophyll levels (typically <1.5 mg m−3) [Arrigo et al., 1998; Marrari et al., 2006] and low rates of primary production [Barbini et al., 2003; El-Sayed, 2005]. Oceanographic research over the last decade revealed that low chlorophyll-a concentrations and low productivity in nutrient-rich surface waters of the SO result from a deficiency of iron [de Baar et al., 1995, 2005; Hopkinson et al., 2007; Martin and Fitzwater, 1988; Wulff and Wängberg, 2004]. Iron fertilization experiments in the SO have clearly demonstrated that iron enrichment increased algal photosynthetic competence, quantum yields of photosynthesis, growth rates, chlorophyll biomass, nitrate uptake, and altered algal community structure toward dominance of diatoms [Boyd, 2002; Coale et al., 2004; Gervais et al., 2002; Hiscock et al., 2008; Hopkinson et al., 2007; Olson et al., 2000; Petrou et al., 2011]. Most iron enrichment experiments in the SO have been conducted in mid or late austral summer, but the physiological status of phytoplankton at the beginning in early summer has scarcely been reported.
[3] Mesoscale patches of high chlorophyll concentration/productivity often occur in frontal regions [Froneman et al., 2001; Holm-Hansen et al., 2004; Laubscher et al., 1993], coastal/neritic waters [Arrigo and McClain, 1994; Holm-Hansen and Mitchell, 1991], the Marginal Ice Zone [Froneman et al., 2001; Sullivan et al., 1993; Westwood et al., 2010], and polynyas [Arrigo and van Dijken, 2003; Moore and Abbott, 2000; Smith and Gordon, 1997]. These “naturally fertilized” waters have been considered as most productive and biologically important areas of the SO [Blain et al., 2007; Mongin et al., 2008]. Blain et al. [2007] suggested that the efficiency of natural iron fertilization from below the euphotic zone to the surface may be 10 times higher than the previous estimates obtained from short-term artificial iron-addition experiments. Recently, Park et al. [2010] argued that the dynamics of natural iron fertilization in the southwest Atlantic sector of the SO is driven by topographic effects that are responsible for a weak seasonality of the phytoplankton bloom and a strong interannual variability in that region.
[4] Iron plays a key role in the photosynthetic electron transport in algal cells [Raven et al., 1999]. Development and improvement of physiological or chemical diagnostics of nutrient stress and, in particular, of iron limitation have been a continuing goal of the oceanographic community [Hopkinson et al., 2007]. Active fluorescence techniques have been used to probe the function of photosynthetic light harvesting, photosystem II (PSII) photochemistry, and electron transport [Kolber et al., 1988]. The fluorescence yield from PSII is highly variable and dependent on the physiological state of phytoplankton, and is affected by environmental factors, including light and nutrient availability [Falkowski and Kolber, 1995]. Fast repetition rate (FRR) fluorometry is a nondestructive and rapid method, and it has been used to monitor variations in the quantum efficiency of PSII (Fv/Fm) and the functional absorption cross-section of PSII (σPSII) [Falkowski and Kolber, 1995; Kolber and Falkowski, 1993]. Variable fluorescence signals are particularly sensitive to iron limitation [Greene et al., 1992; Vassiliev et al., 1995] and can be used to monitor alterations in the physiological state of phytoplankton in response to iron stress [Falkowski et al., 2004].
[5] Variable fluorescence measurements can be used to assess iron limitation without conducting incubations [Olson et al., 2000]. However, the paucity of studies of phytoplankton physiology in relation to iron supply is attributed to the difficult accessibility of the SO. A recent study showed that phytoplankton iron stress is an important control on photosynthetic physiology in the southern Drake Passage [Hopkinson et al., 2007]. Their incubation experiments suggested that mixing between iron-limited Antarctic circumpolar current (ACC) waters and iron-replete shelf waters in the southern Drake Passage may stimulate high phytoplankton biomass in the areas downstream of the mixing region. Dulaiova et al. [2009] also emphasized the importance of lateral mixing, which provides ample iron supply to account for the biological demand of dissolved iron to the offshore regions of the Drake Passage. The authors found that lateral fluxes of iron overwhelm vertical inputs and vertical export from the water column, but quantitative contributions of these fluxes to the regulation of iron supply in the upper water column requires further investigation.
[6] Here, we report results of a multidisciplinary study in the Drake Passage and the western part of the Scotia Sea in the early austral summer of 2001 (30 November to 9 December), conducted as a part of the Korea Antarctic Research Program (KARF 15th). The Drake Passage is located south of South America, and the ACC flow is the strongest in this area. The Polar Front is located circa 57 to 63°S in the Drake Passage [Moore et al., 1999], and this area is known to be a typical high nutrient low chlorophyll (HNLC) region. However, in the shelf zone and near the Antarctic Peninsula, high chlorophyll concentrations seasonally appear due to the effects of water inflow from the Bellingshausen Sea and the Weddell Sea. Weak seasonality and strong interannual variations of phytoplankton blooms prevailed in the Scotia Sea, except for the western part of the Drake Passage region [Park et al., 2010]. The main entrainment of upwelled Upper Circumpolar Deep Water, which can cause a large iron flux into the surface layer, occurs in autumn and winter [Hoppema et al., 2003]. Smith et al. [2000] suggested that phytoplankton biomass and primary productivity in the Ross Sea was likely limited by low irradiance in the early spring, but was limited by trace-metal availability in the late spring and summer. If low irradiance, but not iron deficiency, would be a primary limiting factor in the Scotia Sea in early summer, one would expect no physiological signatures of iron limitation at that time.
[7] The objective of our field study was to elucidate the natural variability in phytoplankton photosynthetic characteristics as an index of iron limitation in the western part of the Scotia Sea in early summer and to relate this variability to physicochemical conditions of the water column. We report here the physiological state of phytoplankton and environmental conditions in the western Scotia Sea before and at the beginning of a phytoplankton bloom.
2. Methods
2.1. Observations
[8] The cruise was conducted in a region located between 52–63°S and 66–52°W and included the Drake Passage, the western Scotia Sea, the northeastern Antarctic Peninsula, and the Bransfield Strait (Figure 1 and Table 1) in the early austral summer of 2001 aboard R/V Yuzhmorgeologiya. Stations (23 total) were located across three major fronts: the Polar Front (PF), the Southern Antarctic circumpolar current front (SACCF), and the Southern Boundary (SB) of the ACC. Figure 1 shows the monthly composite surface chlorophyll-a image from sea-viewing wide field-of-view sensor (SeaWiFS) level-3 data (version 2010.0), together with bottom topography (1000, 2000, and 3000 m isobaths) and major fronts. The Shackleton Fracture Zone is also depicted as a major topographic barrier for the ACC in the region.
| Station | Latitude (°S) | Longitude (°W) | Local Date | Local Time |
|---|---|---|---|---|
| Ws2 | 52.98 | 52.05 | 30 Nov | 20:30 |
| WsA | 53.51 | 52.00 | 1 Dec | 06:31 |
| Ws3 | 54.03 | 51.94 | 1 Dec | 10:50 |
| WsB | 54.49 | 52.01 | 1 Dec | 16:39 |
| Ws4 | 54.97 | 51.98 | 2 Dec | 22:07 |
| WsC | 55.49 | 51.99 | 2 Dec | 05:28 |
| Ws5 | 55.97 | 52.01 | 2 Dec | 10:38 |
| WsD | 56.45 | 51.98 | 2 Dec | 17:19 |
| WsPF | 56.22 | 52.00 | 2 Dec | 20:26 |
| Ws6 | 56.97 | 52.00 | 3 Dec | 04:32 |
| WsE | 57.52 | 52.01 | 3 Dec | 13:08 |
| Ws7 | 59.01 | 52.08 | 4 Dec | 01:26 |
| Ws8 | 59.99 | 51.99 | 4 Dec | 14:00 |
| Sf1 | 57.95 | 58.99 | 5 Dec | 14:40 |
| Sf2 | 58.98 | 59.04 | 6 Dec | 02:25 |
| Sf3 | 60.50 | 59.02 | 6 Dec | 15:11 |
| Kr1 | 61.17 | 57.79 | 6 Dec | 23:01 |
| Kr2 | 61.47 | 57.62 | 7 Dec | 2:40 |
| Kr4 | 62.11 | 57.13 | 7 Dec | 10:04 |
| Kr6 | 61.78 | 56.33 | 7 Dec | 21:30 |
| Kr8 | 61.13 | 56.76 | 8 Dec | 03:03 |
| Kast1 | 61.69 | 55.02 | 8 Dec | 11:44 |
| Kast2 | 62.25 | 57.61 | 8 Dec | 15:10 |
[9] Temperature, salinity, and underwater light intensity were measured using a Sea-Bird conductivity, temperature, and depth probe (CTD) equipped with a submersible photosynthetically available radiation (PAR) sensor (Biospherical Instruments Inc. QSP-200L). Water samples were taken from standard depths (surface, 10, 20, 30, 50, 75, 100, and 150 m, respectively) using 10 L Niskin bottles mounted on the CTD-rosette. Chlorophyll-a concentrations were measured following the fluorometric method [Parsons et al., 1984] with a Turner fluorometer (TD-700 model) on board the ship. The instrument was calibrated against a commercial chlorophyll-a standard (Turner Designs Inc.). We filtered 500 mL of water through Whatmann GF/F filter that was then soaked in 90% acetone for 24 h prior to absorption measurements. Two size fractions of phytoplankton, nanoplankton (<20 µm), and microplankton (>20 µm) were gently separated by serial filtration (without vacuum pressure). Then, nanofraction chlorophyll-a concentrations were measured in the same way as bulk populations.
[10] Water samples for nutrient analysis were poured directly from Niskin bottles into acid cleaned polyethylene bottles, and the samples were maintained in a freezer (<−20°C) until analysis. Nutrient concentrations were measured using an automatic ion analyzer (LaChat model, QuikChem AE) and samples were validated against a standard seawater sample (Wako Pure Chemical Industries, CSK (Cooperative Study of the Kuroshio and Adjacent Regions) standard solution).
[11] For analysis of the phytoplankton community structure, a CTD-Niskin rosette sampler was also used to take water samples from 10 m depth. Aliquots of 125 mL were preserved with glutaraldehyde (1% final concentration). Sample volumes of 100 mL were filtered through Gelman filter paper (0.45 µm pore size, 25 mm diameter). The filters were mounted on microscope slides with water-soluble embedding medium (HPMA, N-(2-hydroxypropyl)methacrylamide) on board [Crumpton, 1987; Kang et al., 2001]. Then, HPMA slides were examined using light microscopy (LM) and epifluorescence microscopy (Nikon type 104). Cells were counted in random fields until 300–500 had been observed in total. Most Phaeocystis cells present were of the solitary form. Solitary Phaeocystis cells were distinguished from other autotrophic flagellates based on cell size (3–5 µm) and shape, and the presence of flagella. To estimate phytoplankton carbon biomass, biovolume was calculated based on the size of each cell using an image analyzer with a camera mounted on the microscope. Biovolume was then converted to biomass according to Menden-Deuer and Lessard [2000].
2.2. Determination of Photochemical Parameters
[12] A Chelsea Instruments FASTtracka FRR fluorometer was used to measure variable fluorescence of phytoplankton from discrete samples. After collection from Niskin bottles, samples were kept at in situ temperature in amber glass bottles for 30 min for low-light adaptation and recovery from nonphotochemical quenching. Test measurements of dark fluorescence recovery showed that longer adaptation did not affect, within ∼5% precision, the fluorescence readings, suggesting that a 30 min interval was sufficient for full recovery. Analysis of vertical profiles of FRRF fluorescence showed that Fv/Fm values recorded in surface samples were as high as those deeper in the mixed layer, also suggesting full recovery from photoinhibition and nonphotochemical quenching.
[13] The FRR fluorometer was programmed to generate a saturating Single Turnover Flash consisting of 100 subsaturating flashlets (each 1.1 µs long followed at a 50% duty cycle). PSII parameters such as the minimal fluorescence yield (F0; when all reaction centers are open), the maximal fluorescence yield (Fm; all reaction centers are closed), the functional (or effective) absorption cross-section of PSII (σPSII) were measured as described in Kolber et al. [1988]. The Chelsea FRS program (version 1.8) was used for postprocessing of acquired data. The Fv/Fm was calculated as a ratio of variable fluorescence (Fv = Fm − F0) to the maximum one (Fm). Fluorescence measurements were corrected for the blank signal recorded from filtered seawater as described in Bibby et al. [2008]. We made 10 replicates and then calculated the average value and standard deviation.
2.3. Physical Environments
[14] Under conditions of light winds and solar warming, SST variations of 1 or 2°C can occur within the first 1–2 m near the surface [Price et al., 1986]. Therefore, we chose 10 m as our reference depth to avoid diurnal cycles in the top few meters of the water column. The mixed layer depth (MLD) was defined as the depth at which the potential density is equivalent to that at 10 m and a 0.2°C decrease in temperature from 10 m depth. This definition has been widely employed in previous studies [Breugem et al., 2008].
3. Results
3.1. Hydrography and Chlorophyll Distributions
[16] The SeaWiFS chlorophyll-a image of the study region for December of 2001, together with station locations, is shown in Figure 1. During the cruise, surface chlorophyll-a concentrations (determined on water samples) were relatively low in the ACC region (0.16–0.36 mg m−3), but high in other regions (0.52–2.06 mg m−3). A major transition in surface chlorophyll concentrations occurred near the SACCF, with the shift toward high chlorophyll waters in the south. High chlorophyll concentrations near the Sf3 station in the satellite image seems to be the bloom developed after the cruise (i.e., in mid or late December). A time series of SeaWiFS chlorophyll maps of the study area (not shown) revealed that the sampling period was just preceding the spring bloom in the western Scotia Sea. The bloom peaked in mid-January of 2002, i.e. a month after the cruise.
[17] Figure 2 shows profiles of temperature, salinity and chlorophyll-a concentration measured from the water samples in the upper water column at 23 stations. Generally, the MLDs were >50 m, with the exception of Ws5, WsD, Ws6, Sf1, Kr1, and Kr2 stations located near the fronts. MLD ranged from 16 (WsD) to 110 m (Kr6). The uniform water column structure in the center of the PF during the cruise was unexpected. However, subsurface chlorophyll maximums (SCMs) were clearly seen in WsPF, Ws6, Sf1, Sf3, and Kr1 stations. In addition, a distinct chlorophyll maximum at the surface was observed at Ws7, Ws8, Kr4, Kr6, and Kast2 stations. Generally, the width of the Antarctic PF is about 43 km [Moore et al., 1999] and the PF is defined by sea surface temperature of around 1.5°C or by the northern limit of the temperature minimum layer. In this study, the surface temperature at the PF station was somewhat higher (2.77°C; probably it was not exactly in center of the PF), but the temperature minimum layer (ca., 80–140 m) was repeatedly observed at the stations south off the PF (Ws6, WsE, Ws7, Sf1, Sf2, Sf3, and Kr1 stations) [see also Holm-Hansen et al. 2004].
[18] For the purpose of comparison, the study area was divided into three regions, based on the hydrographic characteristics [Moore et al., 1997; Orsi et al., 1995]. The regions were defined as follows: (1) ACC waters north off PF and PF station, (2) ACC waters south off PF, and (3) shelf and the Southern Boundary of the ACC.
3.2. ACC Waters North Off Polar Front Region and PF Station
[19] Eight stations belong to this region (Ws2, WsA, Ws3, WsB, Ws4, WsC, Ws5, and WsPF). Surface water temperatures were moderately high and ranged from 5.25°C to 2.77°C (north to south). Surface chlorophyll-a concentrations were very low (0.17–0.28 mg m−3). Vertical distributions of Fv/Fm, σPSII, and chlorophyll-a concentration (averaged over the three stations, Ws2, Ws5, and WsPF, with standard deviations) are shown in Figure 3 (left). SCMs were found at ca. 50 m depth at Ws5 and WsPF stations, though MLD was as deep as 93 m at the WsPF. The surface values of Fv/Fm varied from 0.3 at Ws5 and WsPF to 0.42 at Ws2 (Table 2). Values of Fv/Fm increased with depth and reached the maximum of about 0.57 below 100 m. σPSII also increased with depth by SCM, but decreased below SCM.
| Sts. | Region | Temp | MLD | Fo | Fm | Fv/Fm | σPSIIb | TNc | Pc | Sic | Chl-a | <20 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ws2 | North to pf | 5.25 | 55 | 1.100 | 1.920 | 0.426 | 680 | 22.9 | 1.65 | 14.4 | 0.284 | 87 |
| Ws5 | North to pf | 5.31 | 24 | 0.667 | 0.995 | 0.330 | 656 | 23.6 | 1.75 | 17.4 | 0.168 | 93 |
| WsPF | PF | 2.77 | 93 | 1.650 | 2.300 | 0.283 | 852 | 25.6 | 1.85 | 20.9 | 0.240 | 58 |
| Sf1 | South to pf | 0.94 | 41 | 1.790 | 2.400 | 0.253 | 760 | 24.1 | 1.72 | 30.2 | 0.248 | 50 |
| Sf2 | South to pf | 0.93 | 82 | 0.723 | 1.500 | 0.517 | 609 | 28.1 | 2.11 | 63.6 | 0.260 | 37 |
| Sf3 | South to pf | 0.47 | 55 | 0.927 | 1.430 | 0.351 | 540 | 25.7 | 1.87 | 50.0 | 0.360 | 74 |
| Kr4 | Shelf | 0.42 | 70 | 5.280 | 7.830 | 0.325 | 711 | 27.3 | 2.13 | 102.7 | 2.060 | 54 |
| Kr6 | Shelf | 0.40 | 110 | 2.750 | 4.690 | 0.415 | 570 | 27.9 | 2.13 | 99.8 | 1.068 | 40 |
| Kast1 | Shelf | 0.52 | 79 | 1.350 | 2.340 | 0.424 | 508 | 28.5 | 2.19 | 100.2 | 0.708 | 86 |
| Ws7 | SB | 0.79 | 68 | 1.500 | 3.050 | 0.508 | 589 | 28.4 | 2.17 | 79.6 | 0.696 | 66 |
| Ws8 | SB | 0.24 | 82 | 2.440 | 3.820 | 0.361 | 803 | 28.3 | 2.20 | 84.6 | 0.904 | 88 |
- a Temperature (°C), mixed layer depth (MLD, m), FRRF parameters (F0, Fm, Fv/Fm, and σPSII), nitrate + nitrite (TN), phosphate, silicate, chlorophyll-a concentration (mg m−3), and percentage of <20 µm size-fractionated chlorophyll-a (%).
- b Units are A2.
- c All units are µM kg−1.
3.3. ACC Waters South Off Polar Front Region
[20] Three stations in the western Scotia Sea (WsD, Ws6, and WsE) and another three stations in the Drake Passage across the Shackleton Transverse Ridge (Sf1, Sf2, and Sf3) belong to this region. Surface temperatures ranged from 0.47 to 1.64°C and surface chlorophyll-a concentrations were also low (0.17–0.36 mg m−3). Among these stations, we measured FRRF parameters only at three stations, Sf1, Sf2, and Sf3 (Table 2). SCMs were also found at ca. 50 m depth at Sf1 and Sf3 stations and at ca. 100 m at Sf2 station. Fv/Fm value at the surface was high at Sf2 (0.52), but was low at Sf1 (0.25) and Sf3 (0.35). Fv/Fm increased with depth, except Sf2 where Fv/Fm was high throughout the entire water column.
3.4. Shelf and the Southern Boundary of the ACC Region
[21] Seven stations in the shelf around South Shetland Islands (Kr1, Kr2, Kr4, Kr6, Kr8, Kast1, and Kast2) and two stations in the southwestern Scotia Sea (Ws7 and Ws8) were attributed to this region. Surface temperatures ranged from 0.20 (Kr8) to 0.79°C (Ws7) and sea surface chlorophyll-a concentrations were relatively high (∼0.50–2.0 mg m−3). FRRF parameters have been measured at three stations in the shelf region (Kr4, Kr6, and Kast1) and at two stations in the Southern Boundary of the ACC (Ws7 and Ws8) (Figure 3). MLDs were deep (>68 m) and no SCMs were observed. Surface Fv/Fm values ranged from 0.32 to 0.51, and were intermediate between the low values found in ACC waters and the highest values found at Sf2 station (Table 2). At all stations, Fv/Fm increased with depth. σPSII was vertically homogeneous at all stations except Ws7, and this was probably attributed to a weak stratification of the water column in this region.
3.5. Nutrients and Phytoplankton Communities
[22] Vertical distributions of nitrate + nitrite, phosphate, and silicate concentrations in the three regions are shown in Figure 4. All macronutrient concentrations were high even at the surface at all stations (Table 2 and Figure 4). The minimum amount of nitrate + nitrite was observed at Ws2 station (22.9 µM kg−1), and the maximum was 28.5 µM kg−1 at Kast1. The phosphate and silicate concentrations also were lowest at Ws2, and highest at the southern stations (especially silicate concentrations). Vertically, gradients of nutrient concentration increases with depth were much larger in the ACC waters south of Polar Front region than in the ACC waters north of Polar Front region or the shelf and the Southern Boundary of the ACC. A notable feature of the chemical properties was that the major nutrients were saliently high on the surface at Sf2 and were more uniform with depth. In addition, nitrate + nitrite and phosphate concentrations in the Southern Boundary of the ACC (Ws7 and Ws8) were as high as in the shelf, but silicate concentrations were not.
[23] The PF marks the location with strong meridional gradients in both temperature and nutrients (Figure 5). In contrast, low levels of surface chlorophyll-a concentration prevailed not only in the ACC waters north off PF, but also in the ACC waters south off PF and PF station region. The surface chlorophyll-a concentrations were highest in Kr4 (∼2.0 mg m−3), and moderately high in another shelf and the Southern Boundary of the ACC region (0.5–1.0 mg m−3). The nanofraction (<20 µm) of chlorophyll-a accounted for 37 to 96% of total chlorophyll-a concentration, with the smallest values observed in the shelf region. The percentage of nanofraction was >80% in most ACC waters except WsPF, Sf1, and Sf3 stations (Figure 5). The nanofraction of chlorophyll-a was also high at Ws7 and Ws8 stations (86.2% and 95.6%, respectively).
[24] The abundance of three dominant phytoplankton assemblages varied considerably among stations (Figure 6). While the nano size flagellate Phaeocystis sp. was prevalent in the study area, with its maximum abundance of 2650 cells mL−1 observed at the shelf station (Kr4). Phaeocystis sp. also prevailed at Sf2 and Ws7 stations (where Fv/Fmvalues were highest at the surface; Table 2) and at Ws8. Diatom (with chain forming diatoms of the genera Chaetoceros spp. and Fragilariopsis sp.) was abundant in the PF as well as at Sf1 and Sf3. In the shelf and the Southern Boundary of the ACC region, Cryptomonas sp. was abundant, with the maximum of 2740 cells mL−1 observed at Kr4 (the station of highest surface chlorophyll-a concentration). Outside this region, Cryptomonas sp. formed a minor component.
4. Discussion
4.1. Hydrographic Properties
[25] The ACC consists of a number of high-velocity fronts separated by relatively quiescent zones. The meridional position of these Antarctic fronts and the flow of the ACC through the Drake Passage are constrained by local topography [Sprintall, 2003]. Our cruise track crossed over three major fronts, including PF, SACCF, and SB. The dynamics of these fronts exerts a strong influence on phytoplankton distribution and productivity [Laubscher et al., 1993]. Generally, the ACC waters are nutrient impoverished, with localized enrichments driven by upwelling or cross-frontal mixing [Moore and Abbott, 2000]. These nutrient enrichments could promote phytoplankton growth [Moore and Abbott, 2000]. Mixing between the iron-poor Antarctic surface water of the ACC and iron-rich waters from the shelf provides a physical control that could be responsible for enhanced primary productivity in the southwestern Scotia Sea [Zhou et al., 2010]. Moreover, Shackleton Jet, which is the southward detour of ACC branch into the Shackleton Gap between the Shackleton Transverse Ridge and Elephant Island, is accelerated further to 70 cm s−1 [Zhou et al., 2010].
[26] In this study, we classified three regions based on their location relative to the major fronts. To quantify water column properties per se, we further used a WI [Zhou et al., 2010] to disclose the hydrological similarity of a given water to the waters of the ACC or of the shelf (Figure 7). We have chosen Ws2 and Kast1 as standards for the ACC and the shelf waters, respectively. When the value of WI is close to 0, the water T-S characteristics are close to that of the ACC; and when WI is close to 1, the water T-S characteristics are close to that of the shelf, and any water with the value of WI between 0 and 1 could be formed as a percentage mixture between the two water masses [Zhou et al., 2010].
[27] All values of WI were close to zero in the ACC waters north off PF and PF station region, while most values of WI were around 1 on the shelf, except Kr2 (Figure 7). In addition, all values of WI were >1 in the ACC waters south off PF and PF station region, except Sf2, though Sf2 station was located far from the shelf. Striking differences also were observed at Ws7 and Ws8, which were quite different from not only northern offshore station (WsE) but also stations on the shelf. Although this approach per se is not sophisticated for such a broad area, it is worthwhile to be used to distinguish water column properties from a hydrographical point of view.
4.2. Variability of Photochemical Parameters
[28] Variability in Fv/Fm is generally associated with changes in the physiological state of phytoplankton [Olaizola et al., 1996], and inversely correlates with the availability of nutrients, especially iron in the SO [Boyd and Abraham, 2001; Olson et al., 2000] and in the subarctic Pacific Ocean [Suzuki et al., 2002]. During our cruise, the macronutrients were abundant and, thus, were not a limiting factor (Table 2 and Figure 4). Following Martin et al. [1990], many studies have shown that low productivity in the nutrient-rich waters of the eastern equatorial Pacific and the SO results from a deficiency of iron. FRRF-derived PSII characteristics, especially Fv/Fm and σPSII, were found to be sensitive indicators of iron limitation [Greene et al., 1992; Hopkinson et al., 2007; Sosik and Olson, 2002; Vassiliev et al., 1995]. The quantum yields of carbon fixation by phytoplankton in the SO are also Fe driven, with low quantum yields being indicative of iron stress [Hiscock et al., 2008]. In addition, Olson et al. [2000] found significant correlations not only between ambient iron concentrations and Fv/Fm, but also between ambient iron and the response of Fv/Fm to iron enrichment in the SO. Therefore, based on previous lab and field research, we may conclude that the low values of Fv/Fm observed in the upper-layer phytoplankton in our study were indicative of iron stress.
[29] The other factors that may potentially reduce the measured values of photosynthetic efficiency (Fv/Fm) are photoinhibition [Falkowski et al., 2004] and grazing by zooplankton [Fuchs et al., 2002]. Photoinhibition is attributed to the photodamage of PSII reaction centers by supraoptimal irradiance and observed during the day in the near-surface layer. This effect is dynamic [e.g, Falkowski et al., 2004] and usually recovers, within 30–60 min, under low irradiance. Because our measurements were conducted on low-light acclimated samples (see section 2.), the effect of photoinhibition was removed. Grazing by zooplankton may reduce the bulk Fv/Fm values due to a fraction of damaged cells (with low Fv/Fm) in the population and due to the presence of pigment degradation products in water. Because the gross photosynthesis and growth rates of phytoplankton are greatly reduced by low light availability at the bottom of the euphotic zone, the effect of grazing, e.g., the amount of pigment degradation products [Gervais et al., 2002] should be most pronounced below the mixed layer. Our measured vertical profiles of Fv/Fm showed that Fv/Fm remained high below the mixed layer and decreased in the upper part of the water column. This decrease in Fv/Fm cannot be explained by the effects of grazing. Our data corroborate the results of Holm-Hansen et al. [2004] who noted that light availability and grazing have relatively small impact on the spatial variability of phytoplankton biomass of the Scotia Sea, compared to the effects of iron availability.
[30] In the SO, iron may originate from land (continent and islands), melting sea ice [El-Sayed, 2005; Martin et al., 1990; Sedwick et al., 2000; Sullivan et al., 1993], and the bottom or deep iron-rich water masses [Blain et al., 2007]. In shallow coastal waters around the Antarctic Peninsula, high iron concentrations may be due to runoff from the land or release of iron by resuspension of bottom sediments [Ardelan et al., 2010]. Among other sources, melting icebergs can provide substantial inputs of iron [Smith et al., 2007], but icebergs were spatially restricted during our study. In contrast, the ACC waters in the central Scotia Sea are considered to be iron-impoverished. The high phytoplankton biomass in the central Scotia Sea might only be speculated as the result of natural iron fertilization by waters from the peninsular area [Ardelan et al., 2010]. In our study area, aeolian inputs of iron are also negligible because of the long distance from the continent or big islands. Moreover, during our study, sea-ice had already retreated to the south (as was evident from remote sensing data), and all stations were far away from the marginal ice zone. Therefore, we could expect that the main sources of iron supply to the upper water column were mixing and upwelling.
[31] From the results of iron enrichment incubations, Hopkinson et al. [2007] concluded that iron was not a limiting factor in shelf waters in the southern Drake Passage region. Mixing of a small amount of high-iron shelf waters with severely iron-depleted ACC waters is a plausible mechanism to maintain elevated phytoplankton biomass found downstream in the ACC. Hopkinson et al. [2007] hypothesized that phytoplankton blooms were the result of natural iron additions to ACC waters. Dulaiova et al. [2009] also reported that lateral iron supply were significantly higher than vertical upward fluxes in the southern Drake Passage. However, their study site was located south of the SACCF (northern region of Kr8; Figure 1), which was comparable to our stations Sf2 and Sf3. Iron deficiency may limit phytoplankton growth in pelagic Antarctic waters, but not in coastal waters close to the continent.
[32] Decadal scaled ocean color images show that higher chlorophyll-a concentrations occur mostly to the south of the PF and to the north of the SB, except for the Antarctic Peninsula shelf [Park et al., 2010]. Also, high chlorophyll-a concentrations around the South Shetland Islands extend as a band toward the northeast over the deep waters of the central portion of the Scotia Sea [Ardelan et al., 2010]. These data imply that natural iron fertilization may enhance primary production in the regions of topographic effects rather than in the vicinity of iron-rich waters.
[33] While surface Fv/Fm values did not covary with chlorophyll-a concentrations during our study, there was a spatial pattern of variability of Fv/Fm in relation to the topographic effects. For instance, the Fv/Fm values at the surface were high at Sf2 and Ws7, where nutrient concentrations were also higher than in the vicinity, though the surface chlorophyll-a concentrations were not (Table 2). The Sf2 station was located on the right side of Shackleton Transverse Ridge, which acts as a topographic barrier (against eastward flow of the ACC) capable of promoting vertical mixing [Park et al., 2010]. In case of Ws7, oceanic advection could bring a favorable nutritional condition through mixing. Shackleton Jet may stimulate mixing between the ACC and iron-replete shelf waters [Zhou et al., 2010]. In this context, the mixing processes may act as natural iron fertilization of downstream waters and be responsible for the elevated chlorophyll concentrations or quantum efficiencies of photosynthesis, Fv/Fm (e.g., at Ws7).
[34] Vertical transport of iron through upwelling or water mass intrusion may also bring new sources of iron to surface waters. Peloquin and Smith [2007] suggested that episodic intrusions of modified circumpolar deep waters from off the continental shelf into the euphotic zone in the southern Ross Sea could have supplied sufficient amount of iron to fuel diatom bloom in the summer. From their study in the Atlantic sector of the Southern Ocean, Klunder et al. [2011] also pointed out that vertical mixing and upwelling were the most important mechanisms of iron supply to the upper layer. However, Hiscock et al. [2003] noted that because the newly upwelled, iron-rich upper circumpolar deep water is only present at the surface north of the SB, the waters south of this front are iron poor. This may explain our observed low values of Fv/Fm (0.36) at Ws8, though chlorophyll-a concentration was elevated (0.90 mg m−3).
[35] The maximum FRRF-derived values of Fv/Fm in nutrient-replete eukaryotic algae, including diatoms, are ∼0.65 [Kolber and Falkowski, 1993]. The maximum values of Fv/Fm observed in iron-enriched waters of the SO are ∼0.55 [Coale et al., 2004; Gervais et al., 2002]. In this study, the average Fv/Fm value was 0.38, thereby Fv/Fm varied from ∼0.25 to ∼0.43 in the ACC waters, and from ∼0.32 to 0.42 in the shelf and the Southern Boundary of the ACC region, except for two stations with high Fv/Fm (∼0.51). Our measured Fv/Fm values are, on average, lower than the previous observations in the SO. Peloquin and Smith [2007] reported Fv/Fm of 0.43 (average) in the Ross Sea, and they found exceptionally high Fv/Fm values (0.5–0.65) for most of the seasonal phytoplankton bloom. Fragoso and Smith [2012] also observed relatively high value (mostly >0.4) in all blooms in the Ross Sea and the Amundsen Sea, with the maximum values >0.6. However, these observations were conducted in the austral summer and were obtained with a pulse amplitude modulation (PAM) fluorometer, which usually overestimates Fv/Fm readings by ∼20%, compared with the FRRF. Although Sedwick et al. [2000] noted that iron concentrations in surface waters were relatively high at the beginning of the bloom mainly due to deep mixing and sea-ice melting, our results suggest that this may not be the case for the whole Scotia Sea. Our field measurements revealed physiological signatures of iron limitation in the western Scotia Sea even in the early austral summer.
4.3. Fv/Fm Related to Phytoplankton Communities
[36] Generally, larger size assemblages dominate nutrient-replete waters and phytoplankton blooms, such as the Antarctic PF and marginal ice zone [Bracher et al., 1999]. In contrast, nanophytoplankton and picophytoplankton tend to dominate open waters of the SO [Froneman et al., 2001; Laubscher et al., 1993; Weber and El-Sayed, 1987]. Iron enrichment stimulates a shift in the community structure toward larger diatoms (reviewed by de Baar et al. [2005]). Suzuki et al. [2002] also observed a dramatic floristic shift from phytoflagellates to diatoms during an iron enrichment experiment in the subarctic Pacific Ocean. The phytoplankton crop size on the shelf exhibits a seasonal change, with microplankton being dominant in December but nanoplankton dominant in February and March [Holm-Hansen and Mitchell, 1991]. However, in our study, diatoms were dominant only at Sf3, and Phaeocystis species prevailed (Figure 6). Olson et al. [2000] observed that, in the Ross Sea and the PF during late summer, cryptophytes almost always had high Fv/Fm and showed only minor responses to iron addition. In contrast, during our study, Fv/Fm was not high in the areas where Cryptomonas species predominated (Kr4 and Kr6), (Table 2). Although small cells, such as nanoflagellates and picoeukaryotes, better sequester trace elements such as iron [Sunda and Huntsman, 1997], there is an evidence that smaller cells may also be iron limited, although to a lesser extent [Boyd and Abraham, 2001; Olson et al., 2000]. Phaeocystis antarctica (a haptophyte with colonial and single-celled morphologies; predominant species in the Antarctic Sea) apparently has a greater demand for iron than diatoms [Sedwick et al., 2007]. Moreover, P. antarctica often dominates mixing zones because it has a better ability to maintain near-maximal photosynthetic rates at much lower irradiance levels than diatom do [Arrigo et al., 1999]. Therefore, our observations can be explained by the biochemical characteristics of Phaeocystis. During our study, Phaeocystis species were dominant at Sf2, Kr4, Ws7, and Ws8, where MLDs were deep (82, 70, 68, and 82 m, respectively; Table 2 and Figure 6).
[37] The majority of biological programs in the SO were conducted in the austral summer, from late December to February. In contrast, our study was conducted much earlier in the austral summer when iron limitation was expected to be less critical. However, our data suggest strong signatures of iron limitation of phytoplankton photophysiology at that season. On the contrary, the observed high values of Fv/Fm (>0.5) in near-surface phytoplankton at Sf2 and Ws7 stations suggest the importance of topographic effects related to hydrography. Although there are no reliable iron data, our results imply that not only light intensity but also iron availability is an important factor for phytoplankton growth phase in the study area.
Acknowledgments
[38] We thank the members of 15th Korean Antarctic Research Program onboard the R/V Yuzhmorgeologiya who collaborated on this research. This study was supported by KOPRI project PP13020. MYG was, in part, supported by the NASA Ocean Biology and Biogeochemistry Program. We also thank two anonymous reviewers and the Editor for their help in strengthening the manuscript.
