Introduction

The brown alga Fucus vesiculosus L. is mainly a marine, North Atlantic, intertidal species (Powell, 1963), but the alga also grows in the sublittoral of the brackish Baltic Sea, along a salinity gradient down to the range of 2–4 psu (Wærn, 1952). Before the recent discovery of Fucus radicans L. Bergström et L. Kautsky (Bergström et al., 2005), F. vesiculosus was thought to be the only large belt-forming algal species in the Bothnian Sea, a northerly part of the Baltic Sea (Wallentinus, 1979). F. radicans is most certainly endemic to the Baltic Sea, grows side-by-side with F. vesiculosus and was earlier thought to be the same species.

The environmental factors differ greatly between the Norwegian Sea (part of the Atlantic Ocean) and the Bothnian Sea (part of the Baltic Sea). In contrast to the marine tidal Norwegian Sea (34 practical salinity units, psu), the Bothnian Sea has low salinity (4–5 psu), no tides and periodically an ice-cover. Environmental factors fluctuate rapidly for algae in the tidal zone of the Norwegian Sea. The differences in the local environments between the Norwegian Sea and the Bothnian Sea have given rise to differences in morphology as well as different physiological acclimation and/or adaptation capacities and mechanisms in the marine Norwegian Sea and the brackish Bothnian Sea ecotypes of F. vesiculosus. The brackish ecotype of F. vesiculosus is smaller with a thinner thallus (Kalvas & Kautsky, 1993), has greater respiration rates, lower maximum photosynthetic capacities (P max), lower mannitol concentration, lower tolerance of emersion stress and contains fewer water soluble organic compounds than the marine ecotype of F. vesiculosus (Bäck et al., 1992; Ruuskanen & Bäck, 1999; Pearson et al., 2000; Nygård & Ekelund, 2006; Gylle et al., 2009a, b).

An important protein, involved in linear electron transport in oxygenic photosynthesis, is the chlorophyll (Chl) a-binding core protein D1 (PsbA, 37 kDa) in the photosystem II (PSII) reaction centre. The D1 protein is rapidly cycled during illumination and disruption of D1 protein cycling or losses of D1 protein pools are central to the photoinhibition of photosynthesis (Nixon et al., 2005; reviewed by Nixon et al., 2010). One of the important proteins in the photosystem I (PSI) reaction centre is the Chl a binding core protein, PsaA (82–83 kDa; Hall & Rao 1999; reviewed by Santabarbara et al. 2010). PSI is involved in linear electron transport as well as cyclic electron transfer around PSI. The energy generated in the electron transport chain is stored in ATP and the reducing power is stored in NADPH. ATP and NADPH are used in the photosynthetic rate-limiting step of fixation of CO2 by ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco). The most common structure of Rubisco consists of eight large (L) subunits (50–55 kDa) and eight small (S) subunits (15 kDa), L8S8 (~550 kDa; Raven, 1997).

P max is lower in the Baltic Sea ecotype compared to the Atlantic Ocean ecotype of F. vesiculosus and if the Baltic Sea ecotype is kept for 1 week or longer in higher salinity (10–35 psu) the P max increases considerably (Bäck et al., 1992; Nygård & Ekelund, 2006; Nygård & Dring, 2008). The reason for the higher P max in the Atlantic Ocean ecotype of F. vesiculosus relative to the Baltic Sea ecotype, and for the increase of the P max in Baltic Sea ecotype of F. vesiculosus at higher salinity, have been suggested to be either an increase in the relative amounts of Rubisco, or an up-regulation of its activity (Nygård & Dring, 2008).

The aim of this study was therefore to investigate photosynthetic differences between the marine, Norwegian Sea ecotype and the brackish, Bothnian Sea ecotype of F. vesiculosus and to see whether the photosynthetic differences could be connected with the relative amounts of D1 protein (PSII), PsaA (PSI) protein and/or Rubisco.

Fucus radicans was also included in this study in order to investigate if the native Bothnian Sea species shows other adaptations to the environment than in the original marine F. vesiculosus ecotype in the Bothnian Sea. The proteins have not been evaluated in absolute terms and therefore the amounts of D1 (PSII), PsaA (PSI) and Rubisco are compared among the algae on a relative basis. Further to this, the aim was to see whether the photosynthetic differences could be connected with the relative amounts of D1 protein, (PSII), PsaA (PSI) protein and/or Rubisco when the brackish Bothnian Sea ecotype of F. vesiculosus was affected by varying salinities (5, 10, 20 and 35 psu) in the short-term.

Materials and methods

Algal material

Fucus vesiculosus samples were collected from the intertidal area of the Norwegian Sea at Trondhjem Biological Station (Norway; 63°43′N; 10°39′E; 34 psu) and from 2 to 4 m depth from the Bothnian Sea on the East side of Åstön (Sweden; 62°24′N; 17°45′E; 4–5 psu). F. radicans was collected from the Bothnian Sea at the same time and place as F. vesiculosus. The samples were collected in May 2006 and the water temperature was 6–8°C. The algae were transported in their natural water and darkness to the laboratory. Five individual samples of F. vesiculosus from the Bothnian Sea were prepared for an experiment as described below. For the aim of measuring the relative amounts of photosynthetic proteins from the collected algae groups, epiphytes were removed and tissue from the vegetative parts of the algal tips from ten individuals of each ecotype of F. vesiculosus and from ten individuals of F. radicans were immediately ground separately with a mortar and pestle in liquid N2 and frozen at −80°C until use. To establish the electron transport rates (ETR) in the three algal groups, fresh thallus from five individuals of each ecotype of F. vesiculosus and from five individuals of F. radicans were collected at the same places as above in February to March, 2010. The water temperature was 0–4°C. ETR was measured within 24 h after arrival to the laboratory and the samples were kept in darkness at 6°C until measurements.

Acclimation to changed salinity

For each experiment, an individual of F. vesiculosus from the Bothnian Sea was cleaned from epiphytes (approximately 10–15 cm long parts of thallus were used) and separated into five samples with two to four branches (depending on size) per sample. One of the samples was placed in Bothnian seawater (BSW, normal salinity 4–5 psu) and four samples in 5 psu artificial seawater (ASW), each in an individual 7 L aquarium. This was repeated with five individuals, so that in the salinity experiment each assay was done on five samples from different individuals. ASW was a mixture of deionised water, sea salt (Instant Ocean, United Pet Group, Cincinnati, OH, USA), inorganic compounds (0.344 mg L−1 NH4Cl, 2.05 mg L−1 NaNO3 and 0.016 mg L−1 NaH2PO4) and bicarbonate (1 mM NaHCO3, Sigma-Aldrich, St. Louis, MO, USA). The concentrations of inorganic compounds were chosen to be similar to those in BSW, according to an analysis done by LMI AB, Helsingborg, Sweden, and the concentration of bicarbonate was the same as in BSW (Raven & Samuelsson, 1988). Before changing salinity, algae were allowed to acclimate for 3 days to the irradiance, temperature and basic growth medium. Thereafter, samples incubated in ASW were placed in salinities of 5, 10, 20 or 35 psu; the sample in BSW was kept as a control. The growth medium was changed every 2.5 days and after 1 week of treatment, tissue from the vegetative parts of the algal tips (approximately three cm of each tip) was ground in liquid N2 and frozen at −80°C until use. During the whole experiment, the algae were kept in 12-h dark-light rhythm at the irradiance of 90 μmol photons m−2 s−1 (Philips Master TL-D 36 W/840, Amsterdam, the Netherlands; measured with LI-COR LI-189 quantum sensor, Lincoln, NE, USA) at 6–8°C.

Electron transport

For the measurements of the ETR from the five individuals of each ecotype of F. vesiculosus and F. radicans, tissues from the vegetative parts of the algal tips were used. ETR was estimated from Chl a fluorescence measurements by rapid light curves using a pulse amplitude modulated fluorometer (Diving PAM, Walz, Effeltrich, Germany). After an initial dark-adaptation of 20 min, nine steps with increasing irradiance between 0 and 700 μmol photons m−2 s−1 (30 s at each step) were used to induce photosynthesis, interrupted by dark periods of 2 min. The temperature was kept approximately at 8°C during the measurements. ETR was measured at three different points at each thallus tip (n = 5) and the mean of the three measurements was used as each individual ETR. The ETR was calculated as:

$$ {\text{ETR }} = {\text{PAR}} \times{\text{AF}}\times\Updelta F/F^{\prime}_{\text{m}}\times0. 5 $$

AF (the absorption factor) is the fraction of incident PAR absorbed by the thalli (plant tips, n = 5) and was estimated according to the method of Beer et al. (2000) and Nygård & Dring (2008). The AF obtained was 0.83 for F. vesiculosus from the Norwegian Sea, 0.78 for F. vesiculosus from the Bothnian Sea and 0.80 for F. radicans. ΔF/F′m is the effective quantum yield measured in light. PAR was multiplied by 0.5 to allow for an equal involvement of the two photosystems (Beer et al., 2000; Nygård & Dring, 2008).

The x axis of the rapid light curve is presented as absorbed light (PAR × AF × 0.5) and the PSII photochemical efficiency was measured as the maximum quantum yield (Y 0) as recommended by Saroussi & Beer (2007). Y 0 can also be expressed as F v/F m where F v is the variable fluorescence, difference between the maximum (F m) and minimum fluorescence, and F m is the maximum fluorescence reached on dark-adapted algae when all PSII reaction centres are closed (Krause & Weis, 1991). An investigation by Gylle et al. (2011) indicated that an assumption of equal involvement of the two photosystems in marine and brackish ecotype of F. vesiculosus and F. radicans is an incorrect simplification. However, more studies have to be done before the exact involvement of PSI and PSII in the absolute ETR have been totally clarified. The P max was calculated with Photosyn Assistant software (Ver. 1.1.2, Dundee Scientific, Dundee, Scotland, UK).

Net photosynthesis measured by oxygen evolution

When the Bothnian Sea ecotype of F. vesiculosus had been treated for 1 week in BSW, 5, 10, 20 or 35 psu, net photosynthesis was measured as oxygen evolution with an oxygen electrode (Rank Brothers Ltd, DM 10, Cambridge, England). Measurements were performed in a saturated inorganic carbon environment (3 mM NaHCO3), using the vegetative tips of the thallus (pre-cut 24 h before measurement). The sample was placed in a vertical position above a magnetic stirrer and facing the light projector. Light periods (5 min) with increasing irradiance in eight steps between 0 and 300 μmol photons m−2 s−1 were used, interrupted by dark periods of 5 min. In the salinity-treated Bothnian Sea ecotype of F. vesiculosus, efficiency of use of the absorbed light in photosynthesis at limiting irradiance (α) was calculated from the light-limited initial slope of the photosynthesis/irradiance curve (P/I curve) between three data points from 0 to 35 μmol photons m−2 s−1. P max was calculated with Photosyn Assistant software.

Protein extraction

For immunoblotting, proteins from 75 mg of tissue were extracted in 1.5 mL of buffer according to a protocol modified from Li et al. (1998). The buffer (pH 7.4) contained 100 mM sodium ascorbate (KEBO Lab, Stockholm, Sweden), 50 mM sodium 3-[N-morpholino] propane sulfonic acid (MOPS), 3 mM EDTA, 5 mM dithiothreitol (DTT), 2% polyvinylpolypyrrolidone (PVPP), 12 mM 3-[(3-cholamidopropyl) dimethyl-ammonio]1-propanesulfonate (CHAPS), 2% Triton X-100 and 10 μL protease Inhibitor Cocktail (Sigma-Aldrich). Aliquots of the protein extracts were stored at −80°C until SDS-PAGE and immunoblotting analysis. For spectrophotometric determination of Chl a and Chl c (c 1 and c 2) concentrations, 90% acetone was used and the absorbance was measured at 630, 664 and 750 nm (Varian Cary 50 Scan, Agilent Technologies, Santa Clara, CA, USA; Porra, 1991).

SDS-PAGE and immunoblotting

Proteins in the solubilised samples containing 5 μg Chl a were used for the detection of D1 and PsaA proteins and 1 μg Chl a for the detection of Rubisco. The proteins were electrophoretically separated in a 15% (w/v) polyacrylamide resolving gel containing 6 M Urea, 0.66 M Tris (pH 8.8) and a 5% (w/v) polyacrylamide stacking gel-containing 0.125 M Tris (pH 6.8) using the discontinuous Laemmli buffer system (Laemmli, 1970). Electrophoresis was performed at room temperature in a Mini-Protean II apparatus (Bio-Rad, Hercules, CA USA) for 1 h at 3 mA. Protein standards (Bio-Rad, item number 161-0304) and protein extracts for all treatments were separated by SDS-PAGE and transferred electrophoretically to 5% (w/v) BSA blocked nitrocellulose membranes (Bio-Rad) and probed using the PsbA polyclonal total IgY antibody (anti-D1, PSII), 1:50,000 dilution; PsaA polyclonal antibody (anti-PSI-A), 1:500 dilution and global RbcL polyclonal antibody (anti-Rubisco), 1:5,000 dilution (Agrisera, Umeå, Sweden). After incubation with the secondary antibody IgY, conjugated to peroxidise, for PsbA (Sigma-Aldrich, 1:10,000 dilution); IgG, conjugated to horseradish peroxidase, for PsaA and Rubisco (Agrisera, 1:50,000 dilution and Abcam, Cambridge, MA, USA, 1:20 000 dilution, respectively), the antibody-protein complexes were visualised using ECL chemiluminescence detection reagents (GE Healthcare, Waukesha, WI, USA) and developed on X-OMAT film (VWR, East Grinstead, United Kingdom). To obtain sharp bands, the optimal concentrations of Chl a and antibodies were determined in preliminary tests. The immunoblotting was repeated 5–7 times for each replicate and for each class of protein and the developed photos were analysed by using ImageJ software (developed by Rasband (1997–2011), National Institutes of Health, Bethesda, MD, USA; use reference, Shim et al., 2010).

Data analysis

To analyse the differences, a general linear model analysis of variance (GLM) or one-way ANOVA, followed by Tukey’s pairwise comparison, were performed. All factors were categorical and fixed. Tukey’s pairwise comparison was performed for a complete comparison of all parts included in the experiments. The confidence intervals were set to 95%. The standard deviation values are given in the figures and tables.

Results

Photosynthesis and relative amounts of D1, PsaA and Rubisco proteins in F. vesiculosus and F. radicans

To analyse the photosynthesis and to see if photosynthetic differences were connected with the relative amounts of D1, PsaA and/or Rubisco in the Fucus strains, analyses of P/I curves (by ETR) and immunoblotting were performed (Figs. 1, 2; Table 1).

Fig. 1
figure 1

Photosynthesis/irradiance curve measured by electron transport in marine F. vesiculosus from the Norwegian Sea (35 practical salinity units, psu) and brackish F. vesiculosus and F. radicans from the Bothnian Sea (4–5 psu). The measurements were completed after an initial dark-adaptation of 20 min and nine steps with increasing irradiance between 0 and 700 μmol photons m−2 s−1 were used (points with incident irradiance above 250 μmol photons m−2 s−1 are not shown). Data represent means of five replicates ± standard deviation

Full size image
Fig. 2
figure 2

The relative amounts of a D1 protein in photosystem II (PSII). b PsaA protein in PSI and c ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in marine F. vesiculosus from the Norwegian Sea (35 practical salinity units, psu) and brackish F. vesiculosus and F. radicans from the Bothnian Sea (4–5 psu) were determined by immunoblotting and by ImageJ software. Data were represent means of ten replicates blotted 5–7 times each (for each protein) ± standard deviation. Different letters indicate significant (P < 0.05) difference between the Fucus strains and lack of letter means no differences compared to the other Fucus strains. The y axis presents the results in percent relative to the marine F. vesiculosus

Full size image
Table 1 Photosynthetic maximum capacity (P max) measured by electron transport, the efficiency to use absorbed light in photosynthesis (maximum quantum yield, Y 0) and the relative D1/PsaA ratio in marine F. vesiculosus from the Norwegian sea (35 practical salinity units, psu) and brackish F. vesiculosus and F. radicans from the Bothnian Sea (4–5 psu)
Full size table

The results confirm that P max was significantly higher (57%) in the marine ecotype of F. vesiculosus compared to the brackish ecotype of F. vesiculosus (P < 0.01; Fig. 1; Table 1). The P max was also significantly higher (51%) in the marine ecotype of F. vesiculosus compared to F. radicans (P < 0.01; Fig. 1; Table 1). No differences between the brackish Fucus species were detected. The results did not indicate any differences in Y 0 between the Fucus strains (Table 1).

The results from the estimations of the relative amounts of the D1 protein on a Chl a basis did not indicate any significant differences between the two ecotypes of F. vesiculosus or between the two brackish Fucus species (Fig. 2a). Between the marine ecotype of F. vesiculosus and F. radicans, however, a significant difference could be detected with a greater relative amount of D1 protein in F. vesiculosus (P < 0.01; Fig. 2a). The measurements of PsaA protein showed that all three Fucus strains contained similar amounts of PsaA on a Chl a basis (Fig. 2b). No significant differences in the relative PSII/PSI ratios were detected between any of the Fucus strains (Table 1). The results from the immunoblotting of Rubisco indicated that the relative amount of Rubisco in both strains of F. vesiculosus was almost double that in F. radicans (P < 0.001; Fig. 2c).

Effect of salinity on net photosynthesis and the relative amounts of D1, PsaA and Rubisco proteins in the brackish, Bothnian Sea ecotype of F. vesiculosus

To test if changes in photosynthetic variables were connected with the relative amounts of D1, PsaA and Rubisco when the brackish ecotype of F. vesiculosus was affected by salinity, the alga was treated in a range of salinities and analyses of P/I curves (by O2 evolution) and immunoblotting were performed (Figs. 3, 4; Table 2).

Fig. 3
figure 3

Photosynthesis/irradiance curve measured by oxygen evolution in brackish F. vesiculosus from the Bothnian Sea (4–5 practical salinity units, psu) treated for 1 week in Bothnian seawater (BSW) or in artificial sea water at different salinities (5, 10, 20 or 35 psu). The light (90 μmol photons m−2 s−1):dark period was 12:12 h. The measurements were completed after an initial dark-adaptation of 20 min and eight steps with increasing irradiance between 0 and 300 μmol photons m−2 s−1 were used. Data were represent means of five replicates ± standard deviation

Full size image
Fig. 4
figure 4

The relative amounts of a D1 protein in photosystem II (PSII). b PsaA protein in PSI and c ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in brackish F. vesiculosus from the Bothnian Sea (4–5 practical salinity units; psu) were determined by immunoblotting and by ImageJ software. The algae were treated for 1 week in Bothnian seawater (BSW) or in artificial sea water at different salinities (5, 10, 20 or 35 psu). The light (90 μmol photons m−2 s−1):dark period was 12:12 h. Data represent means of five replicates blotted 5–7 times each (for each protein) ± standard deviation. Different letters indicate significant (P < 0.05) difference between the treatments and lack of letter means no differences compared to the other treatments. The y axis presents the results in percent relative to the algae treated in BSW

Full size image
Table 2 Photosynthetic maximum capacity (P max) measured by oxygen evolution (nmol O2 s−1 g−1 FW), the efficiency to use absorbed light in photosynthesis at limiting irradiance (α: calculated between 0 and 35 μmol photons m−2 s−1) and the relative D1/PsaA ratio in the brackish Bothnian Sea (4–5 practical salinity units, psu) F. vesiculosus after 1 week treatment in Bothnian seawater (BSW) or in artificial seawater at different salinities (ASW of 5, 10, 20 or 35 psu)
Full size table

Short-term salinity treatments had an effect on P max. The algae treated at salinities of 10 and 20 psu had significantly higher P max values than those at BSW or 5 psu (P < 0.01) and the algae treated at 35 psu had significantly higher P max values than those at BSW (P < 0.05; Fig. 3; Table 2). The algae treated at 10 psu had also significantly higher P max than algae treated in 35 psu (P < 0.05; for details in significant differences in P max between the treatments see Table 2). Photosynthetic efficiency (α) was largely unaffected but it was significantly lower in algae treated in BSW compared to algae treated in 20 psu (Fig. 3; Table 2).

The immunoblotting did not indicate any effects of salinity on the relative amounts of D1 protein or Rubisco on a Chl a basis in the brackish ecotype of F. vesiculosus (Fig. 4a, c). In contrast, the relative amount of PsaA protein in PSI on a Chl a basis showed significantly greater amounts of PsaA in algae treated in 10 psu compared to algae treated in BSW and in 35 psu (P < 0.05; Fig. 4b). The PSII/PSI ratio, measured by immunoblot signals of D1/PsaA, was significantly lower in algae treated in 10 psu (0.45) than in algae treated in 35 psu (0.72; P < 0.05; Table 2).

Discussion

Photosynthesis and relative amounts of D1, PsaA and Rubisco proteins in F. vesiculosus and F. radicans

The results indicated a higher P max in the marine ecotype of F. vesiculosus than in the brackish ecotype of F. vesiculosus and F. radicans (Fig. 1; Table 1). The lower P max in the brackish water Fucus than in the marine ecotype of F. vesiculosus is in line with earlier investigations (Bäck et al., 1992; Nygård & Ekelund, 2006; Nygård & Dring, 2008). In this study we made a first attempt to reveal what causes the differences in P max between the Fucus strains by immunoblotting of D1, PsaA and Rubisco. Immunoblot signals in the detections of D1 and PsaA showed a similar D1/PsaA ratio in all three Fucus strains. The results suggest, however, an uneven photosystem protein stoichiometry in Fucus, with more of the PSI protein PsaA relative to the PSII protein D1 (Table 1). This difference between PSI and PSII appears to be in agreement with 77 K Chl a fluorescence emission spectra for the marine F. vesiculosus (PSII/PSI ~ 0.3) and F. radicans (PSII/PSI ~ 0.4; Gylle et al., 2011). 77 K fluorescence emission of the brackish ecotype of F. vesiculosus, however, showed a PSII/PSI ratio ~1 in Gylle et al. (2011). The reason for the higher fluorescence emission in PSII in the brackish F. vesiculosus, than in the other Fucus strains, was concluded to be due to larger light-harvesting antenna of PSII and not due to a truly higher PSII/PSI ratio (Gylle et al., 2011). This interpretation of the 77 K fluorescence emission of the brackish ecotype of F. vesiculosus in Gylle et al. (2011) appears to be in line with the immunoblot signals of D1/PsaA (Table 1). The PSII/PSI ratio differ between different organisms and have been observed to be 1.43–1.72 in some terrestrial plants, 0.43 in blue-green algae and 0.84–1.03 in some green algae (Melis & Brown, 1980; Falkowski et al., 1981; Yamazaki et al., 2005; Fan et al., 2007).

The reasons for generally lower PSII/PSI ratio in marine algae compared to terrestrial plants have been suggested to be: (1) a need of greater ATP/NADPH ratio as energy to import various nutrients and therefore more PSI to mediate cyclic electron transport, and (2) the imbalance in the photosystem stoichiometry protects algae from photoinhibition in the blue–green light that preferentially excites PS II (Yamazaki et al., 2005).

Because of the differences in P max (Fig. 1; Table 1), it was hypothesised that the marine ecotype of F. vesiculosus would have more Rubisco than both the brackish Fucus species. Rubisco is accumulated in parallel with high photochemical activities (Björkman, 1981; Maayana et al., 2008). The results, however, showed approximately the same relative amount of Rubisco in both ecotypes of F. vesiculosus whereas F. radicans had about half the amount (Fig. 2c). These results indicated that the reasons for the lower P max in the brackish ecotype of F. vesiculosus and F. radicans than in the marine ecotype of F. vesiculosus might originate from different underlying causes. Whereas the relatively lower amount of Rubisco may be an important part of the explanation of the low P max in F. radicans, there may be another explanation for the low P max in the brackish ecotype of F. vesiculosus. The difference in P max between the two ecotypes of F. vesiculosus might be related to the difficulties for the algae to adapt to the environment in Bothnian Sea or to normal differences in environmental adaptations. Further investigations are needed in order to understand the low P max in the brackish ecotype of F. vesiculosus and we suggest measuring the rate of CO2 fixation by Rubisco as a first step. Only few investigations have been done on the recently discovered F. radicans and it is important to keep in mind that F. radicans, unlike the brackish ecotype of F. vesiculosus, is a native species in the Bothnian Sea environment. It is interesting to note the similarity in P max between the Bothnian Sea Fucus species despite lower relative amount of Rubisco in F. radicans pointing towards a more efficient CO2 fixation in F. radicans. A speculation worth take in consideration is that the brackish ecotype of F. vesiculosus compared to the marine ecotype and F. radicans might have a less efficient C4-like pathway to accumulate CO2. This may increase a need of higher amount of Rubisco, relative to the P max, in the brackish ecotype of F. vesiculosus. Wang et al. (2011) showed higher abundance and activity of Rubisco in gametophytes compared to sporophytes and assumed that a more marked C4-like pathway supply the sporophytes with CO2 and made it unnecessary to have a high abundance of Rubisco.

Effect of salinity on net photosynthesis and relative amounts of D1, PsaA and Rubisco of the brackish, Bothnian Sea, ecotype of F. vesiculosus

In general, there were higher P max values in the brackish ecotype of F. vesiculosus treated in the higher salinities, 10, 20 and 35 psu, compared to P max in algae treated in BSW and 5 psu. The most favourable salinity was 10 psu followed by 20 psu (Table 2). These results were in line with earlier ETR measurements when the alga was treated in different salinities together with low or high concentrations of dissolved inorganic carbon and nutrients (Nygård & Dring, 2008).

In this first attempt to explain the differences in the P max of the brackish F. vesiculosus treated in different salinities we did not observe any differences in the relative amount of either D1 or Rubisco (Fig. 4a, c). The relative amount of PsaA protein (Fig. 4b) and the D1/PsaA ratio (Table 2), however, indicated a salinity effect. The relative amount of PsaA was greatest in algae treated in 10 psu and it was significantly greater in 10 psu compared to algae treated in BSW and 35 psu. The greater amount of PsaA in algae treated in 10 psu was in line with the photosynthetic measurements (Fig. 3; Table 2). The explanation might be that the algae need to produce more ATP, and greater amount of PsaA may increase the possibility to have a greater flow of cyclic electron transport around PSI to serve a higher rate of CO2 concentrating mechanisms and/or a higher rate of CO2 fixation by Rubisco. It has earlier been confirmed that greater relative amounts of PSI and greater cyclic electron flow around PSI might occur in conditions with an increased need for ATP (Mandori & Melis, 1984; Anderson et al., 1995; Tanaka et al., 1997; Hall & Rao, 1999; Yamazaki et al., 2005). Gao et al. (2011) have also showed that the electron flow around PSI was significantly elevated in Ulva sp. during desiccation which indicates that PSI-driven cyclic electron flow might provide some protection of the photosynthetic apparatus and extra ATP under stress conditions. As a consequence of more PsaA, the D1/PsaA ratio was lowest in algae treated in 10 psu and significantly lower relative to the algae treated in 35 psu (Table 2). Earlier studies have confirmed a change in PSII/PSI ratio as a response to CO2 supply, irradiance, nitrogen supply and because of salinity shock (Mandori & Melis, 1984; Kim et al., 1993; Berges et al., 1996; Tanaka et al., 1997). It is tempting to speculate that an increased amount of ATP might be used to serve a more efficient C4-like pathway and thereby an accumulation of CO2.

The indicated lower D1/PsaA ratio in algae treated in 10 psu in Table 2 was not in line with 77 K fluorescence emission spectra (Gylle et al., 2011) and in order to understand what causes the differences between the studies we suggest an investigation of the light-harvesting antenna proteins as a first step.