The latitudinal variation in the photosynthesis-irradiance (P-E) relationship of phytoplankton was investigated across a cyclonic eddy in the Antarctic Divergence zone along the 140°E longitude. Both the maximum photosynthetic rate (P*max) and maximum light utilization coefficient (α*) were lower at stations within the eddy as compared with those in the continental shelf-slope area and the offshore area. The low-density water mass of the eddy prevented the upward and horizontal supply of macronutrients and micronutrients, which likely resulted in a reduction of P*max and α*. In contrast, both P*max and α* were large and variable in the slope area, probably owing to a local stability resulting from the less-saline surface water. The present study indicates the possibility that longitudinal variations in the latitudinal pattern of meanders, eddies, seasonal sea-ice extent, and topographical conditions significantly influence the mesoscale variations in the P-E parameters and primary production in the Southern Ocean.
 In the Southern Ocean, satellite ocean color remote sensing has revealed the complicated latitudinal distributions of chlorophyll a (Chl a) along various fronts of the Antarctic Circumpolar Currents (ACC) [e.g., Comiso et al., 1993; Moore et al., 1999; Moore and Abbott, 2002]. Moreover, eddies and meanders around the Antarctic Polar Front (APF) and Antarctic Divergence (AD) zone further complicate the distributions of Chl a, inducing vertical mixing of the water column and upwelling of nutrients from deeper water [Moore et al., 1999; Moore and Abbott, 2002; Hirawake et al., 2003]. Hirawake et al.  observed patchy, eddy-like distributions of Chl a in the AD zone around 140°E on many occasions during the 2001/2002 austral summer season on satellite ocean color images. The existence of this eddy was confirmed by satellite sea surface height anomaly data and in situ hydrological observations [Aoki et al., 2007].
 Temperature, light, trace metals, and physical mixing processes have all been proposed as key factors governing phytoplankton production in the Southern Ocean [cf. Holm-Hansen et al., 1977; El-Sayed, 1987; Figueiras et al., 1998; Bracher et al., 1999; Hiscock et al., 2003]. Among them, the dependence of photosynthesis on light can be analyzed using the parameters of photosynthesis-irradiance (P-E) relationships that are central for measuring and modeling phytoplankton photosynthesis and for evaluating the physiological states of populations. However, compared with primary production, information on the P-E parameters, such as the maximum photosynthetic rate (P*max) and maximum light utilization coefficient (α*), have been reported for only a limited region of the Southern Ocean, i.e., the Antarctic Peninsula region [e.g., Figueiras et al., 1998]. In particular, regional variations in the P-E parameters across eddies or meanders occurring in the fronts are poorly understood.
 The present study aims to describe the latitudinal variation in the P-E parameters to examine whether the photosynthetic activity of phytoplankton responds to a changing light/nutrient regime across a cyclonic eddy appearing in the AD zone in the Australian sector of the Southern Ocean during austral summer.
2. Materials and Methods
 Field experiments were carried out along the 140°E meridian in the Southern Ocean in February 2002 during the JARE-STAGE (Japanese Antarctic Research Expeditions-Studies on Antarctic Ocean and Global Environment) cruise on board the R/V Tangaroa (Figure 1). Downward PAR was measured with an underwater PAR sensor (QSL, Biospherical) equipped with a CTD sensor. Macronutrients (nitrate, nitrite, ammonium, phosphate, and silicate) were determined using an autoanalyzer (TRAACS 2000, Bran+Luebbe). Chl a was determined fluorometrically after extraction by N, N-dimethylformamide [Suzuki and Ishimaru, 1990]. The Chl a-specific light absorption coefficient (a*ph) was measured by the quantitative filter technique (QFT) method [Kishino et al., 1985].
 Photosynthetic rates were determined at a depth of 10 m (surface) and at the bottom of the euphotic zone (1–10% light depths) based on the uptake of H14CO3− during incubation in a photosynthetron [Lewis and Smith, 1983]. The obtained photosynthetic rates were fitted with the expression of Platt et al. :
where P*(PAR) [mg C (mg Chl a)−1 h−1] is the instantaneous photosynthetic rate normalized to Chl a at a level of PAR, α* [mg C (mg Chl a)−1 h−1] [μ mol m−2 s−1]−1 is the initial slope of the P-E curve at limiting light levels, β* [mg C (mg Chl a)−1 h−1] [μ mol m−2 s−1]−1 is a parameter to characterize photoinhibition, and P*s [mg C (mg Chl a)−1 h−1] is the potential maximum light-saturated rate of photosynthesis under noninhibiting conditions. P*max [mg C (mg Chl a)−1 h−1] and P*s can be related as follows.
 The instantaneous primary production P(z, t) was calculated from the concentration of Chl a, the P-E relationship, the surface PAR records, and vertical profiles of PAR by a wavelength-integrated model [Behrenfeld and Falkowski, 1997b]. The P-E parameters at a depth z were obtained from the linear interpolation of the parameters at a depth of 10 m and at the bottom of the euphotic zone. Chl a at the depth z was similarly calculated using the neighboring depths. The water-column-integrated daily primary production ΣPP was obtained by integrating P(z, t) over the duration of daytime D and over the euphotic layer Zeu.
3. Hydrographic Structure and Chl a Distribution
 On the basis of both hydrographical [Sokolov and Rintoul, 2003] and geological conditions, the sampling sites were divided into four different areas: the continental shelf-slope area (Stns. 6, 6.1, 6.2, 7, 7.1, 7.2, 7.3., and 8), within a cyclonic eddy (Stns. 4 and 5), an offshore area (Stns. 1, 2, 3, and 9), and a subantarctic area (Stn. 10). A water mass with low salinity (33.65–33.76) and low sigma-t (26.96–27.13) existed around 63°S–64°S (Stns. 4 and 5) within the upper 50-m depth (Figure 2). This water mass was repeatedly detected as a cyclonic eddy at the same position during the 2001/2002 spring-summer seasons by satellite ocean color images, sea surface height anomaly data, and in situ hydrological observations [Hirawake et al., 2003; Aoki et al., 2007]. The southernmost station (Stn. 8) was partially covered by sea ice and was called the ice edge zone.
 Macronutrient levels at the surface were generally high over the study areas, except silicate in the subantarctic area (Figure 2). The highest concentration of silicate at the surface was in the slope area (37–56 μM); the concentration was locally minimal at the center of the eddy (<20 μM). In the slope area and at the eddy stations, deeper water with high concentration of macronutrients was lifted upward; however, the low-density water mass within the eddy appeared to prevent the upward transport of macronutrients to the upper 50-m depth. Chl a decreased northward and sharply declined in the slope area. The depth of the euphotic zone (Zeu) became deeper toward the north, ranging from 35 to 97 m. The depth of the upper mixed layer (Zmxl) was shallower than Zeu by 32–53 m at the stations in and around the eddy (Stns. 3, 4, 5, and 6). The differences between Zmxl and Zeu were small at the other stations.
4. Latitudinal Variation in Phytoplankton Photosynthesis
 Both P*max and α* showed a general decreasing trend toward the north (Figure 3). Their values were large and variable in the slope area, relatively low and stable in the offshore area, and locally minimal in the eddy area. The value of P*max ranged from 1.01 to 4.08 mg C (mg Chl a)−1 h−1. The value of α* ranged from 0.005 to 0.051 [mg C (mg Chl a)−1 h−1] [μ mol m−2 s−1]−1. The value of β* was mostly zero or low at both two depths (<0.0005 [mg C (mg Chl a)−1 h−1] [μ mol m−2 s−1]−1). The light saturation index (Ek = P*max/α* [Sakshaug et al., 1997]) tended to be low over the area, ranging from 60 to 216 μ mol m−2 s−1. The value of Ek was higher at a depth of 10 m as compared with the bottom within the eddy, the ice edge, and offshore areas. The difference was particularly large within the eddy area. The value of ΣPP was markedly high in the slope area because of the high photosynthetic activity and phytoplankton biomass (Table 1). In spite of a reduction in the photosynthetic activity within the eddy, ΣPP was larger than in the offshore area because of the larger phytoplankton biomass.
Table 1. Latitudinal Variations in the Water Column Integrated Primary Production, Chl a, Primary Productivity, Daily Air Surface PAR, and Light Utilization Efficiency in February 2002a
Primary production, ΣPP; Chl a, ΣChl; primary productivity, PPB; daily air surface PAR, (Par(+0)); light utilization efficiency, Ψ.
(mg C m−2 d−1).
(mg C [mg Chl a]−1 d−1).
(mol photon m−2 d−1).
(mg C [mg Chl a]−1 m2 [mol photon] −1).
Continental Shelf Slope
 There was a general increasing trend in a*ph toward the north in the slope area at both depths (Figure 3). This latitudinal variation in a*ph may be accounted for by changes in the daily solar radiation and the percentage of diatoms to total phytoplankton biomass (M. Miki et al., unpublished data, 2002). A regional difference in the maximum quantum yield of photosynthesis (ϕc max = 0.0231 α*/a*ph) became larger compared with those in α* and a*ph. The value of ϕc max ranged from 0.008 to 0.125 mol C [mol photon]−1 in the study area. The above values of the P-E parameters in the present study were within the ranges previously reported for the Southern Ocean during austral summer [e.g., Tilzer et al., 1986; Sakshaug and Holm-Hansen, 1986; Figueiras et al., 1998; Hirawake et al., 2000].
 To examine the factors that regulate the latitudinal variations in the P-E parameters, a simple linear regression analysis was performed. The values of P*max and α* at the bottom of the euphotic zone showed significant positive correlations (p < 0.01, n = 13) with the logarithm of Si (r = 0.62 and r = 0.75, respectively). Similar trends were observed for the P*max and α* values at a depth of 10 m, although the relationships among them were weak and non-significant (p > 0.05). This result indicates that the photosynthetic activity of diatoms in the eddy and offshore areas was limited by the low Si concentration. However, the half saturation constant for the Si uptake by diatoms in the Southern Ocean was reported to be 0.7–10 μM [Nelson et al., 2001], and the surface Si was >15 μM, even at the eddy and offshore areas. The limitations due to low quantities of Si appeared to be minor. Although the relationships of P*max and α* at a depth of 10 m with dissolved Fe were weak and insignificant (p > 0.05), the dissolved Fe at a depth of 5 m was the lowest at Stn. 5 (0.19 nM) within the eddy (X. Lai et al., Spatial and temporal distribution of Fe, Ni, Cu and Pb along 140E in the Southern Ocean in 2001/2002 austral summer, submitted to Marine Chemistry, 2007, herinafter referred to Lai et al., submitted manuscript), where P*max and α* were low. Fe affected the growth of the surface population at Stn. 5, as revealed by an in vitro Fe enrichment experiment (M. Miki et al., unpublished data, 2002). The dissolved Fe was the highest at Stn. 8 (0.83 nM) in the ice edge area. Thus, the low concentration of Fe within the eddy might have affected the photosynthesis performance of phytoplankton. The values of P*max and α* showed a negative linear relationship with water temperature; however, these relationships were considered to be spurious correlations. This is because the water temperature showed a negative correlation with Si at both the depths, and P*max is the indicator of the dark reaction of photosynthesis and increases directly with temperature [Eppley, 1972].
Hirawake et al.  reported that P*max was larger in the southern AD zone at around 140°E in the Southern Ocean during austral summer, which agrees with the present study. On the other hand, Bracher et al.  reported that P*max tended to be large and variable in the northern regions (APF and Subantarctic Polar Front (SAPF)) and small in the southern regions (ACC and marginal ice zone (MIZ)) in the Atlantic sector of the Southern Ocean; this latitudinal pattern was opposite to that obtained in the present study. Hiscock et al.  also reported higher values of P*max and the optimum assimilation rate Poptb [Behrenfeld and Falkowski, 1997a] in the northern region of the maximum gradient of Si (ΔSimax) in the SB as compared to the southern stations along the 170°W longitude in the Pacific sector of the Southern Ocean. They suggested that the photosynthetic activity of phytoplankton was limited by the low dissolved Fe in the southern part of ΔSimax. They also speculated that the difference in the latitudinal pattern of phytoplankton production between 170°W and 140°E could be ascribed to the difference in the extent of the MIZ in early summer. Along the 140°E longitude, the southern part of the AD zone and the continental slope area were shallow and the possibility of vertical transport of dissolved Fe was suggested [Sohrin et al., 2000; Lai et al., submitted manuscript]. As the offshore area was deep and the seasonal ice extended up to 62°S, the winter nutrient recharge was dampened. Contrary to these, along the 170°W line, the areas south of ΔSimax and the AD zone were deep and far from the Antarctic coast; therefore there seemed to be low supply of dissolved Fe. With regard to the offshore area, the seasonal ice extended to just south of the APF (near 61°S), and the macronutrients and micronutrients were recharged during winter. In conjunction with the previously obtained results mentioned above, the present result reveals a longitudinal variation in the latitudinal pattern of the P-E parameters and suggests that the longitudinal variation was dependent on hydrographic conditions and resulted in a longitudinal variation in primary production.
5. Reduction of Phytoplankton Photosynthesis Within the Eddy
 A decrease in P*max and α* and large vertical variations in Ek and a*ph were observed within the eddy formed in the AD zone (Figure 3). These spatial variations in the P-E parameters appeared to arise due to horizontal and vertical isolation of the water mass within the eddy from the surrounding water. The low-density water mass of the eddy prevented an upward supply of nutrients from deeper water and horizontal advection of sea ice/water, which possibly resulted in the reduction of photosynthetic activity at both high (P*max) and low light levels (α*). The vertical separation of phytoplankton at the two depth assemblages appeared to be accelerated by the low-density water mass at the surface, which was suggested by the large difference between Zeu and Zmxl, and resulted in the large vertical variations in the light saturation parameter (Ek) and light absorption coefficient (a*ph).
 In contrast to the present result, phytoplankton biomass and primary production were generally enhanced around the eddy and meander in the APF zone, being supported by the upwelling of nutrients from the deeper layers [Moore and Abbott, 2002; Strass et al., 2002]. The eddy observed in the present study had a cyclonic feature [Aoki et al., 2007] that implied the upwelling of deep water. However, unlike the usual cyclonic eddies, the density of the surface water mass within this eddy was lower than that of the surrounding water; the density difference might result in the inhibition of any upward supply of nutrients into the euphotic zone (Figure 2). Hirawake et al.  suggested that the equatorward excursion of the continental shelf is essential for the formation of the eddy and that a bump centered around 63.5°S may have an influence on the structure. Similar eddy-like structures were also reported in the AD zone around 80°E–110°E [Wakatsuchi et al., 1994] and 131°E–144°E [Aoki et al., 2007]. Hirawake et al.  also suggested that these eddies appeared to accelerate exchange between coastal and offshore waters and cause patch distributions of high-Chl a water on the basis of satellite ocean color images. Therefore, the mesoscale variation in the P-E parameters in the slope area was also likely to have been caused by the eddy in the AD zone.
maximum photosynthetic rate, mg C (mg Chl a)−1 h−1
maximum light utilization coefficient, [mg C (mg Chl a)−1 h−1] [μ mol m−2 s−1]−1
chlorophyll a specific light absorption coefficient, m2 (mg Chl a)−1
maximum quantum yield of photosynthesis, mol C (mol photon)−1
 We thank all the participants of the R/V Tangaroa cruise during the 43rd JARE. Thanks are also due to K. Norisue and Y. Sohrin for the data on dissolved Fe and to M. Fukuchi and T. Odate for conducting the JARE-STAGE cruise. Discussion with B. Griffis was very helpful. T. Y. acknowledges financial support from ANESC, University of Tokyo.