Ultraviolet radiation(UVR)(280-400 nm wavelength)intensity varies seasonally in the Arctic depending upon factors including the extent and spatial distribution of sea ice and the ozone layer(Aas et al., 2002; Sakshaug, 2004; Leu et al., 2006). UVR negatively aff ects intracellular DNA and nutrient uptake(Karentz et al., 1991; Behrenfeld et al., 1995; Boelen et al., 2000; Häder et al., 2014), and inhibits photosynthesis and primary productivity in phytoplankton(Helbling et al., 1992; Häder et al., 2007). Similarly, high levels of UVR can lead to the production of increased levels of reactive oxygen species(ROS)from excited photosynthetic pigments(He and Häder, 2002). However, the eff ects of UVR vary depending on the phytoplankton species composition(Vernet and Whitehead, 1996). These influences can be minimized via photoadaptation or DNA repair mechanisms in phytoplankton(Karentz et al., 1991), or cells can be protected against UVR by UV-absorbing compounds or xanthophyll pigments(Sinha and Häder, 2008).

By producing photoprotective compounds, aquatic organisms protect themselves against strong visible and UV radiation, and can survive in extreme environments(Moeller et al., 2005). Carotenoid pigments function as antioxidants and disperse light energy to prevent intracellular photoinhibition in phytoplankton exposed to excessive light irradiation(Roy, 2000; Laurion et al., 2002). Several types of carotenoids, diadinoxanthin(DD) and diatoxanthin(DT)in particular, function as important xanthophyll pigments in some phytoplankton species, including diatoms, dinoflagellates, and prymnesiophytes(Demers et al., 1991; Arsalane et al., 1994; Moisan et al., 1998; Fujiki and Taguchi, 2001). DT is well known for its photoprotective activity(Demming-Adams and Adams, 1996), and DD has recently been described as a photoprotective compound(Laurion et al., 2002; Demming-Adams and Adams, 2006; Korbee et al., 2010). It has been demonstrated that when algal cells are exposed to high-intensity irradiance, excess energy is reduced by interconversion between DD and DT(Demming-Adams and Adams, 1996; Fujiki and Taguchi, 2001; Dimier et al., 2007).

The prymnesiophyte Phaeocystis spp. is distributed widely in Polar regions, including the Southern Ocean. These species react rapidly to changes in UVR(Karentz and Spero, 1995) and contain high levels of DMSP, the biological precursor of dimethyl sulfi de(DMS)(Keller et al., 1989; Stefels and Van Boekel, 1993; Hefu and Kirst, 1997). Furthermore, a potential function of DMSP as an antioxidant in relieving photo-oxidative stress may be realized during intracellular accumulation of strong light or UV radiation, as well as in the case of nutrient limitation(Sunda et al., 2007; Archer et al., 2009). A previous study showed that the intracellular accumulation of DMSP was influenced by the intensity of UVR in several types of phytoplankton cultured indoors(Archer et al., 2009). In the prymnesiophyte Emiliania huxleyi, the intracellular DMSP concentration was increased by 10%-25% following short exposure to UVR(<1 d)(Slezak and Herndl, 2003). In some cases, the DMSP concentration was increased by 100%, compared to treatment with photosynthetically active radiation(PAR)alone(Sunda et al., 2002).

Mycosporine-like amino acids(MAAs)consist of aminocyclohexenon or aminocycloheximine rings with an amino alcohol substituent(Karentz, 2001). MAA absorbance occurs at wavelengths of 310- 362 nm, and these compounds protect intracellular organs against UVR(Karentz, 2001; Whitehead et al., 2001; Volkmann and Gorbushina, 2006). MAAs are known to have a photoprotective function in aquatic organisms, including phytoplankton, in the case of exposure to excess natural light(Häder et al., 1998). They likely act as a passive type of sunscreen, located in the cytoplasm of the cell(Garcia-Pichel and Castenholz, 1993).

In Kongsfjorden in the sub-Arctic, the quantity of light increases gradually from the end of February, and daylight nights last from April to August. Hanelt et al.(2001)reported that visible radiation(400-700 nm)is present at a maximum of 170 W/m² daily between June and July of 1998, whereas UVR peaked at 16.8 W/m² . The objective of this study was to investigate the distribution of MAAs, xanthophyll pigments such as DD and DT, and -DMSP in a natural phytoplankton community continually exposed to UVR in Kongsfjorden, the sub-Arctic. Levels of these compounds were compared among phytoplankton assemblages to evaluate their respective roles in the photoprotective strategies of these species.

Sampling was conducted in near shore and off shore waters near Kongsfjorden Bay, a glacial fi ord on the western side of Spitsbergen Isl and , Svalbard(79°N and 12°E)from May 22, 2009 to May 29, 2009 during the cruise of R/V FARM(Fig. 1). Surface seawater samples were collected by bucket at 23 sampling stations for the measurements of xanthophyll pigment in phytoplankton, and -DMSP. The collected surface water was fi ltered(1 L; triplicate)separately for each compound(xanthophyll pigment and mycosporine-like amino acids)using pre-combusted(450°C)GF/F filter papers. Samples in aluminum foil were then transported to the laboratory using liquid nitrogen canisters. The 23 sampling stations were classifi ed into 3 regions, the outer(station; C1-C4, A5-A7, and B1-B2), the inner Kongsfjorden Bay(station; K01-K04, T01-T05, and A02-A03) and the off shore region(Fig. 1 and Table 1).

Fig. 1 Locations of sampling stations in the inner and outer regions Kongsfjorden Bay, Svalbard

Figure options

Table 1 Surface seawater temperature(SST), chl a concentration and relative abundance of accessory pigments to chl a ateach station in the inner and outer Kongsfjorden Bay, and off shore Svalbard(SST data from Ha et al., 2012)

Table options

During the cruise of the present study, the UV data(average UV intensity: 13.2 W/m²)were provided by the Alfred Wegener Institute for Polar and Marine Research(AWI). However, there was a daily variation in UVR, which depends upon atmospheric conditions, representing a day and night cycle(Fig. 2). The ocean color images of Moderate Resolution Imaging Spectroradiometer(Aqua MODIS satellite)confi rmed the increase in phytoplankton biomass off shore(Ha et al., 2012). Phytoplankton species composition, including identifi cation and quantitative analysis, was determined at fi ve stations(B01, B03, B09, A05, and T05)(Ha et al., 2012).

Fig. 2 Daily fluctuation of UVR measured spectroradiometrically in air during the samplingperiod from May 22, to May 29 2009 in Ny Ǻlesund

Figure options

Surface seawater was fi ltered over pre-combusted GF/F fi lters(Whatman GF/F fi lters; 47 mm)for the pigment composition of the phytoplankton. The fi lter papers were wrapped in an aluminum foil in order to prevent photolysis when preserved in a deep freezer(-80°C)at the laboratory before further analysis. The fi lter paper was placed in a Teflon bottle with 100% acetone(3 mL)for pigment extraction and 50 μL(1 mg/mL)apo-8-carotennoate(an internal st and ard)was added before extraction. Pigment extracts were sonicated by ultrasonicator(30 s, 50 W; Ulsso Hi-tech ULH-700s) and stored for 24 h at 4°C until analysis. The extract was fi ltered 1 mL by a syringe fi lter(PTFE 0.2 μm Hydrophobic)to remove the debris. The quantitative and qualitative analysis of pigment was performed using High-Performance Liquid Chromatography(HPLC).

The HPLC method for accessory pigment analysis was employed from Zapata et al.(2000). The pigment compounds were separated using a column(Waters symmetry C₈ column(150 mm×4.6 mm, 3.5 μm))with mobile phase A mixed methanol(50%), acetonitrile(25%), and aqueous pyridine solution(25%) and mobile phase B mixed methanol(20%), acetonitrile(60%), and acetone(20%). A segmented linear gradient was programmed as follows(time in min, %B): 0 min, 0%; 20 min, 40%; 26 min, 95%; 38 min, 95%; 40 min, 0%. Initial conditions were reestablished by reversed linear gradient(10 min). Flow rate was 1 mL/min and an injection volume of 100 μL of sample. The individual absorbance of each compound was detected at 430 nm(250-750 nm scan). The identifi cation of peaks was based on the retention time of pigment st and ards(DHI water & Environment, Hørsholm, Denmark). Pigment concentrations were calculated on the base of peak areas in the chromatogram and using an equation according to Park(2006). Results of the analyses were processed by an Agilent HPLC 1200 series ChemStation integrator-processor.

The DMS and DMSP analysis method is described by Park and Lee(2008). Samples for DMSP analysis were collected and preserved as described in Kiene and Slezak(2006). We analyzed the total DMSP(dissolved and particulate forms) and dissolved DMSP concentrations in seawater samples. Samples for assessment of the dissolved DMSP were fi ltered(GF/F fi lter)under gravity prior to fi xation. All DMSP samples were fi rst fi xed with 50% H₂SO₄(5 μL addition per mL sample)on board, and transported to the laboratory. 10 N NaOH was added at 0.25 mL per mL sample, so DMS, converted from the DMSP, reacted overnight in the dark. Then DMS was measured by gas chromatography using a flame photometric detector(GC-FPD). St and ard DMS solutions of known concentration prepared by alkaline hydrolysis of DMSP-CI(Tokyo Kasei Inc., Japan)in an amber vial(30 mL), and a gas-tight Teflon cap was used for calibration of the GC-FPD. Gas st and ards with certifi ed mole fractions of a DMS gas st and ard(Scott Specialty, 3410 ppbv DMS)was also utilized calibration for the response of the GC-FPD.

Mean values and st and ard deviations of all samples were obtained with at least 3 replicates. Statistical signifi cance(P <0.05)of all samples was obtained through one way analysis of variance(ANOVA)by Tukey’s test. Using SPSS program¹¹.5 for Windows(SPSS), each correlation coeffi cient for individual MAAs compounds, DMSP, and xanthophyll pigments was obtained.

The average chlorophyll a(chl a)concentration was 0.28(±0.13)μg/L in the inner bay and 0.42(±0.20)μg/L in the outer bay(Fig. 3a). However, in the off shore region, a relatively high chl a concentration area, the chl a concentration was 2.41(±1.44)μg/L; i.e., 7-10-fold higher than in other waters(Table 1). The ratio of fucoxanthin to chlorophyll a(fuco:chl a ratio), which is an index of diatoms, was higher in the off shore waters(Fig. 3b), exhibiting average values of 0.41(±0.09)in the off shore waters and 0.27(±0.09) and 0.24(±0.14)in the inner and outer bays, respectively. These results indicate that diatoms(Thalassiosira spp.)were the dominant species in the off shore waters. The ratio of alloxanthin to chlorophyll a(allo:chl a ratio)was low at all locations, with the minimum values off shore(Fig. 3c and Table 1).

Fig. 3 Spatial distribution of chl a and accessory pigmentsa. chl a(μg/L); b. fuco/chl a ratio; c. allo/chl a ratio in the waters aroundKongsfjord, west Svalbard.

Figure options

The spatial distribution of chlorophyll-specifi c xanthophylls(xanthophyll cycle compounds); i.e., the DD+DT:chl a ratio, showed a contrasting pattern to the fuco:chl a ratio. The DD+DT:chl a ratio was lower in off shore waters(0.09(±0.02))than in the inner bay(0.34(±0.24)). In particular, the highest ratio of DD+DT:chl a was detected at station A03(0.88), at the entrance of Kongsfjorden Bay(Table 2 and Fig. 4b).

Table 2 The surface seawater ratio of photo-protective pigment(DD+DT), de-epoxy ratio, the ratio of DMSP, and the ratio of total MAAs to chl a at each study site in the inner and outer Kongsfjorden Bay, and off shore Svalbard(MAAs to chl a modifi ed from Ha et al., 2012)

Table options

Fig. 4 Spatial distribution of(a)ratio of total MAAs(TMAAs)(μg/μg chl), (b)ratio of xanthophyll pigments(DD+DT)(μg/μg chl) and (c)ratio of DMSP(μg/μg chl)in the waters around Kongsfjord, west Svalbard The data of total MAAs is modifi ed from Ha et al., 2012.

Figure options

The total ratio of MAAs to chl a was highest in the inner bay(53.79(±7.19)). In the outer bay and off shore waters, the MAAs:chl a ratios were 22.65(±3.31) and 6.28(±0.46), respectively(Fig. 5 and Table 2). The data of total of MAAs had been derived from Ha et al.(2012).

Fig. 5 Comparisons between individual MAAs(μg/μg chl)vs ratio of xanthophyll pigment(DD+DT)(μg/μg chl) Total MAAs(R = 0.902, P <0.01, n = 24), shinorine(R = 0.844, P <0.01, n = 24), palythine(R = 0.883, P <0.01, n = 24), mycosporineglycine(R = 0.925, P <0.01, n = 24). SH: shinorine; PA: palythine; MG: mycsoporine-glycine. The data of individual MAA modifi ed from Ha et al., 2012.

Figure options

The -DMSP:chl a ratio was distributed evenly throughout Kongsfjorden(Fig. 4c), displaying an average of 83.83(±20.41)in the inner bay and 107.89(±40.16)in the outer bay. The off shore waters exhibited the lowest -DMSP content(67.06(±56.51))(Table 2).

The fuco:chl a ratio was markedly higher in the off shore waters than the inner and outer bays, similar to the chl a concentration(Fig. 3a, b). The off shore area is directly influenced by the high-temperature high-salinity current that flows in the Arctic Ocean, and can be distinguished based on the temperature of the surface water(Table 1; from Ha et al., 2012). In by diatoms, such as Thalassiosira antarctica var. borealis, Cylindrotheca closterium and Chaetoceros spp.(Eilertsen et al., 1989; Hop et al., 2002). In the springtime, ~60 species of phytoplankton, mostly diatoms, can be found in the outer waters of Kongsfjorden Bay. According to a previous report, the haptophytes Phaeocystis pouchetii, and the diatom species Chaetoceros socialis and Thalassiosira nordenskioeldii, are dominant in the region of Kongsfjorden Bay for 5 months(Hop et al., 2002). The diff erences between the off shore waters and the inner bay with respect to phytoplankton communities may be due to the influence of the West Spitsbergen Current(WSC)(Hop et al., 2002; Svendsen et al., 2002; Wiktor and Wojciechowska, 2005). Oceanographic conditions in Kongsfjorden, which is located on the west coast of Svalbard at 79°N 12°E, are aff ected by the WSC, which is relatively warm and salty from the northernmost extension of the warm North Atlantic Current(Svendsen et al., 2002). The area west of the shelf is therefore essentially icefree during winter(Aagaard et al., 1987; Gascard et al., 1995). The glacier- and snow-melt water increase to reduce the water transparency in early June; this is an important feature in the inner bay of the Kongsfjorden shelf(Hop et al., 2002). Although the euphotic layer of the inner bay of the fjord can be as shallow as 0.3 m, the euphotic zone in the central part of the fjord can vary between 6 and 25 m because water transparency during the summer is largely dependent on water currents and tides(Svendsen et al., 2002). The allo:chl a ratio showed a distribution that was markedly diff erent from those of the fuco:chl a ratio and the chl a concentration, and the allo:chl a ratio was higher in the inner bay than in the outer bay(Fig. 3c). Alloxanthin is considered a marker pigment of Cryptophyta(Jeffrey and Wright, 1997), which are likely associated with melting glaciers(Lizotte et al., 1996). In addition, Cryptophyta were reported previously in Kongsfjorden by Wiktor(1999) and Okolodkov et al.(2000).

Pigments are recognized for their roles in photosynthesis and photoprotection under UV stress(Jahnke, 1999; Estevez et al., 2001; Zudaire and Roy, 2001; Han et al., 2003). Carotenoids can eff ectively quench ROS(Asada, 1994) and can scavenge radicals(Rijstenbil, 2002; Banaszak, 2003). In dinoflagellates, the de-epoxidation of DD to DT prevents photodamage via heat dissipation(Young and Frank, 1996) and ROS scavenging(Obertegger et al., 2011). Xanthophyll pigments, such as DD and DT, are present in Prymnesiophyceae, Dinophyceae and Bacillariophyceae(Silva et al., 2007; Laviale and Neveux, 2011), and are thought to influence the spatial distribution of these organisms, as does the WSC in the inner bay. The Cryptophyte Rhodomonas baltica does not produce DT or DD but is known to utilize alloxanthin as a xanthophyll pigment(Laviale and Neveux, 2011).

DD and DT are involved in both photoprotection and photoacclimation(Meyer et al., 2000; Brunet and Lavoud, 2010), which are two diff erent(although connected)processes. Photoprotection includes rapid processes(time scale of minutes)such as the deepoxidation of DD(DD-cycle, which forms DT). A typical index of the xanthophyll cycle activity is the so-called de-epoxy ratio(=DT/DD+DT). The increase in the DD+DT pool with respect to the photosynthetically active pigments(Chl)could be more related to long-term(time scale of days)photoacclimation. We compared the DD+DT pool of the natural populations in diatom- and Phaeocystis spp.-dominated regions to underst and long-term accumulation of xanthophyll pigments in phytoplankton cells in Kongsfjorden(Table 2). Although both DD and DT were detected at some sites in the outer bay and off shore, de-epoxidation(de-epoxy ratio), as an indicator of xanthophyll cycle activity, was higher in the Phaeocystis spp.-dominated inner bay than in other regions(outer bay and off shore)(Table 2). Therefore, Phaeocystis spp., which dominated in the inner bay, likely have xanthophyll pigment concentrations that are higher than in other regions, because of the increased activity of the xanthophyll cycle, which in turn depends on photoacclimation.

The DT+DD:chl a and MAA:chl a ratios were higher in the inner bay and lower in the outer bay and off shore waters(Fig. 4). Phaeocystis spp. are commonly observed in the inner bay and are known to exhibit a high level of UV absorbance(Marchant et al., 1991; Hop et al., 2002). Among Cryptophyta, Rhodomonas baltica has been reported to contain high concentrations of UV-absorbing compounds(maximum absorbance at 310 nm; Llewellyn and Airs, 2010). Trees et al.(2000), who participated in 31 cruises to collect samples for HPLC analysis from¹⁹⁸⁵ to 1995 and an additional cruise in 1998, reported an average ratio of xanthophyll pigments to chl a of 0.35, which is similar to our results, although there is some variation. Llewellyn et al.(2012)also reported the pigment and MAA concentrations along the surface waters of the meridional transect of the Atlantic(52°N to 45°S). The reported pigment concentrations(less than 0.4 μg/L)were similar to our results, but their ratios of MAAs to chl a were signifi cantly lower than those in the present study(79°N), likely due to the diff erent sampling latitudes and phytoplankton communities.

In the inner bay where Phaeocystis spp. dominated, xanthophyll pigments showed higher ratios than in other regions(Fig. 4b)because Phaeocystis spp. exhibited active and rapid xanthophyll cycling, unlike diatoms(Meyer et al., 2000). The light saturation required for optimal cell division diff ers between diatoms and Phaeocystis sp.(Jahnke, 1989; Meyer et al., 2000). In this study, diatoms were represented by the slow xanthophyll acclimation and cycling in the outer bay and off shore waters of Kongsfjorden(Fig. 4). The correlation between xanthophyll pigments and MAAs was higher in the inner bay than off shore, where diatoms dominated. The photoprotective mechanisms of phytoplankton may involve UV-absorbing compounds and xanthophyll pigments, the eff ects of which are complementary(Zudaire and Roy, 2001; Ha et al., 2010). In addition, the xanthophyll pigments and MAAs showed a strong correlation(R² =0.9; P <0.000 1)in the inner bay, but weak correlations in other regions(outer bay: R² =0.12; P = 0.4 and off shore: R² =0.4; P = 0.25)(Fig. 5). Both UV-absorbing compounds and xanthophyll pigments disperse light energy(Demming-Adams and Adams, 1996; Sinha et al., 2003); therefore, these compounds can be expected to be complementary in terms of their photoprotective mechanisms. We suggest that the photoprotective mechanisms of these compounds are complementary in the inner bay, which is dominated by Phaeocystis spp.

We hypothesized that -DMSP might be correlated with UV-absorbing compounds, but no signifi cant correlation was identifi ed(correlation coeffi cient: 0.353; P ≥ 0.05)(Table 3). UVR influences DMSP and DMS metabolism in bacteria(Slezak et al., 2001; Toole et al., 2006), and aff ects the intracellular DMSP concentration in several phytoplankton species(Sunda et al., 2002; Van Rijssel and Buma, 2002; Slezak and Herndl, 2003). Phaeocystis spp. are distributed widely in Kongsfjorden and contain high concentrations of DMSP(Keller et al., 1989; Stefels and Van Boekel, 1993). DMSP might function as an antioxidant, similarly to MAAs and xanthophyll pigments, presenting selective photoprotective strategies as diff erent species-specifi c characteristics(Archer et al., 2009).

Table 3 Correlation coeffi cient for individual MAAs compounds, DMSP, and photo-protective pigments

Table options

In this study, -DMSP showed no relationships with xanthophyll pigments or MAAs. This may be a result of diff erences in the species composition and photoprotective strategies of phytoplankton(Archer et al., 2009). During photosynthesis, oxygen is produced by photosystem II, which increases the internal oxygen concentration and augments the potential for ROS formation, especially under stress conditions(Halliwell, 1987). If DMSP acts as a primary source for ROS scavengers, phytoplankton should respond via increased DMSP synthesis and lysis to form DMS under oxidative stress(Sunda et al., 2002). DMSP, xanthophyll pigments and MAA levels showed similar seasonal changes in the Western English Channel(Archer et al., 2009). However, the phytoplankton in Kongsfjorden, particularly Phaeocystis spp., contain high levels of DMSP, which performs a secondary photoprotective function. However, this high level does not coincide with the spatial distribution of MAAs and xanthophyll pigments. Further investigation is needed to underst and the interactions among -DMSP, MAAs, and xanthophyll pigments.

After comparing the photoprotective compounds in marine phytoplankton cells in the Arctic region, we found that although -DMSP exhibits a photoprotective function, depending on environmental factors and interspecies specifi city, no clear correlation was observed between the distribution of - DMSP and the distributions of MAAs and xanthophyll pigments. Xanthophyll compounds and MAAs were present at higher concentrations in the inner Kongsfjorden Bay. The phytoplankton community showed that xanthophyll pigments and MAAs were complementary and exert a photoprotective eff ect in the phytoplankton of the bay. However, - DMSP was correlated only weakly with xanthophyll pigments and MAAs in terms of the species specifi city of phytoplankton under various environmental conditions. Therefore, the photoprotective function of -DMSP should be investigated further.

We thank M. Klisch and D. P. Häder for providing the MAA-st and ards and Y. N. Kim for providing assistance in laboratory work.