Increased CO₂ modifies the carbon balance and the photosynthetic yield of two common Arctic brown seaweeds: Desmarestia aculeata and Alaria esculenta

The atmospheric concentration of CO₂ has increased by 40 % since pre-industrial times due to anthropogenic activities, causing rapid changes in the Earth’s climate system, and predictions state that the current partial pressure of CO₂ (pCO₂) of 390 μatm will be exceeded more than twice by the year 2100 (IPCC 2013). One of the main consequences of this phenomenon is ocean acidification (OA), which consists of an increase in CO₂ dissolved in the upper layers of the ocean with a concomitant reduction in the pH, as well as a decrease in CO₃ ²⁻ concentrations (Doney et al. 2009). Consequently, the ocean has absorbed more than 30 % of the emitted anthropogenic CO₂ and the pH of ocean surface water has decreased by 0.1 since the beginning of the industrial era (IPCC 2013). Model simulations further predict that the Arctic will experience the greatest acidification within the global ocean, with pH decreasing by 0.45 units in the present century, a change that is amplified by more than 20 % due to freshening and increased carbon dissolution in response to sea ice retreat (Steinacher et al. 2009). Therefore, Arctic marine biota faces a surplus of unprecedented challenges that are beyond what science can document based on available data (Wassmann et al. 2011).

The Arctic coastal environment is characterised by relatively constant and near-freezing water temperatures and strong seasonal variations in light and nutrient availability (Hop et al. 2002). The mid-sublittoral zone at Kongsfjorden (Spitsbergen) is dominated by kelp beds down to at least 10 m depth, which have a dominant role in carbon fluxes at a regional scale, characterised by high biomass areas (up to 21 kg wet mass m⁻²; Hop et al. 2002, 2012). Alaria esculenta is one of these dominating kelps together with Laminaria digitata, Saccharina latissima and the kelp-like species Saccorhiza dermatodea. Other brown seaweeds like Desmarestia aculeata appear as undergrowth species, sometimes forming a separate belt between the mid- and low sublittoral (Wiencke et al. 2004; Hop et al. 2012).

Macroalgae fix inorganic carbon mainly through the enzyme Ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), which can only use CO₂ as inorganic carbon substrate for the carboxylase reaction. Initially, the increase in CO₂ was supposed to favour photosynthesis in macroalgae, taking into account that the majority of their Rubisco is not saturated at present CO₂ concentrations (Raven and Beardall 2003). However, in only few species photosynthesis is dependent on CO₂ passive diffusion (mostly reds), but most seaweeds possess carbon concentrating mechanisms (CCMs), which increase CO₂ concentration around Rubisco (up to 1000 times) via facilitated or active transport of CO₂ and/or HCO₃ ⁻ inside the cell, therefore saturating their photosynthetic carbon demand (Giordano et al. 2005). Thus, photosynthesis of macroalgae with CCMs was supposed not to be affected by the increase in CO₂ (Israel and Hophy 2002); nevertheless, these mechanisms are energetically expensive and their partial deactivation at higher concentration of CO₂ could decrease the energetic demand, as it has been previously shown in different algal species (Gordillo et al. 2001; Hurd et al. 2009; Cornwall et al. 2012). In this way, this energy could be invested in other processes such as assimilation of other nutrients, resulting in an increase in the growth rate of those species (Gordillo et al. 2001), although some of them do not show any deactivation of CCMs in response to increased CO₂ (Zou and Gao 2009; Zou et al. 2011a).

A rise in CO₂ concentration does not always have a positive consequence in macroalgae, as this process causes a decreased saturation state of CaCO₃ which could make calcification more difficult for marine calcifying seaweeds (Hall-Spencer et al. 2008), but there are also reports of a negative effect of increased CO₂ on growth rate in non-calcifying ones (Mercado et al. 1999; Israel and Hophy 2002; Gutow et al. 2014), probably due to an inability to compensate carbon fluxes with the metabolism of other nutrients or due to a negative effect on the physiology caused by the decrease in the external pH.

Since the effect of OA on growth and photosynthetic performance has shown to be species specific, with some species benefitting and some others showing inhibition or no response, it is supposed that OA might promote changes at the community level (Olischläger and Wiencke 2013).

The knowledge about the effects of a rise in CO₂ in the carbon budget of a thallus is scarce, with most of the studies focusing mainly on inorganic carbon acquisition, photosynthetic performance and growth, but less attention has been given to carbon losses due to respiration and organic carbon release or carbon accumulation in storage biomolecules as another carbon sink, hence, closing the carbon balance. Previous evidence of a decrease in respiration rate by high CO₂ concentrations has been reported in vascular plants (Bunce and Caulfield 1991; Azcón-Bieto et al. 1994) and in the chlorophyte Ulva rigida (Gordillo et al. 2001), in the latter allowing an increase in growth rate despite the unchanged photosynthetic rate, while other macroalgal species did not change their respiration rate when exposed to increased CO₂ (Zou and Gao 2009; Zou et al. 2011a; Suárez-Álvarez et al. 2012).

Additionally, the release of dissolved organic carbon (DOC) is usually considered one of the main carbon and energy losses in algae, together with respiration and grazing, and has been frequently used to explain the uncoupling between assimilated carbon and biomass in studies about CO₂ effects on algae (Riebesell et al. 2007; Hopkinson et al. 2010). This mechanism is suggested to account for the maintenance of the internal C:N ratio as has been shown for the cyanobacterium Spirulina platensis (Gordillo et al. 1999), for U. rigida (Gordillo et al. 2001) and in coastal phytoplankton (Sobrino et al. 2014).

Despite the high research interest in Arctic ecosystems due to its prime affection by global change, little is known about the consequences of OA on the carbon fluxes in Arctic seaweeds. These organisms commonly dominate the coastal systems, but there is a lack of knowledge about the cellular carbon budget and the competitive advantages and disadvantages triggered by increasing CO₂ levels. Some Arctic species (kelps and other brown seaweeds) have shown to possess nutritional strategies that allow them to cope with long periods of darkness in winter and nutrient depletion in summer, by accumulating photosynthates during summer which support new growth during N-sufficient winter (Chapman and Lindley 1980; Korb and Gerard 2000; Wiencke et al. 2007). Therefore, carbon accumulation in storage biomolecules must be a relevant process in Arctic brown seaweeds, but it is unknown to what extent they are going to be affected by OA. Information on winner and loser species is certainly lacking. This is especially relevant as it is expected that changes in the Arctic macroalgal communities will propagate along the food web.

The aim of this study was to determine the physiological acclimation strategies of two representative species of the Arctic macroalgal community, D. aculeata and A. esculenta, to the increase in CO₂ concentration expected in near-future scenarios. The results provide new highlights on the knowledge of cellular carbon flux, physiology and photochemical performance in Arctic seaweeds as they are affected by OA.

Plant material

Two major Arctic brown seaweeds were examined in this study: D. aculeata (L.) J. V. Lamour (Desmarestiales, Phaeophyceae) and A. esculenta (L.) Greville (Laminariales, Phaeophyceae). The experiments were carried out in July 2013 under laboratory conditions in Kongsfjorden (79°N, 11°E; Spitsbergen, Svalbard; see Fig. 1). The specimens were collected by divers at 5–6 m (D. aculeata) and 8–10 m (A. esculenta) depth in Hansneset and carried immediately to the laboratory in black plastic bags. For A. esculenta, young sporophytes were selected (length: 10–15 cm), but for D. aculeata, only mature sporophytes were available. Thus, we used the whole thallus of A. esculenta, but for D. aculeata we previously cut the apical parts of different thalli, just below the meristematic zone, where the active growth occurs. Thalli were kept in 100 L containers under continuous seawater circulation at around 4 °C for 24 h before the experiments. Visually healthy thalli, free from macroscopic epibiota were chosen for the experiments.

Map of Svalbard archipielago (right) and the Kongsfjord on Spitsbergen (left), with an indication of the collecting site (Hansneset)

Full size image

Experimental setup

Algal specimens were cultured for 7 days at two different dissolved CO₂ concentrations, 390 and 1300 ppm, by aerating the medium with regular air or CO₂-enriched air (final aeration of 1 L min⁻¹). For the high CO₂ treatment, a stream of pure CO₂ was mixed with regular air, yielding CO₂-enriched air of approximately 1300 ppm, and was continuously controlled with a Carbon Dioxide Sensor (AirSense Model 310e, Digital control Systems, Inc., USA). The two CO₂ conditions were verified by measuring seawater pH (NBS-scale) and determining total alkalinity (TA) by potentiometric titrations (Gran 1952) every other day; then, CO₂ speciation was calculated using the CO2calc Package (Robbins et al. 2010), with the CO₂ acidity constants of Mehrbach et al. (1973) and the CO₂ solubility coefficient of Weiss (1974).

Cultures started after 3 days of pre-acclimation to the two conditions to avoid the interference of rapid and transient responses to increased CO₂ levels. As shown in previous studies, a total of 10 days of exposure to different CO₂ concentrations (3 days of pre-incubation + 7 days of incubation) would be enough time for acclimation in seaweeds (Mercado et al. 1999; Andría et al. 2001; Zou 2005). Experiments were carried out in a temperature-controlled room (4 ± 0.5 °C) with a 20:4 h light:dark photoperiod (as a proxy to conditions in the collection site), using 1.5-L perspex cylinders with 0.2-µm-filtered natural seawater (FSW) enriched with nutrients, following a modified, buffer-free recipe of Provasoli (1968). A constant photon fluence rate (PFR) of 25–30 μmol photons m⁻² s⁻¹ at the surface of the FSW in the cylinder was provided by daylight fluorescent lamps (L36W/954 Osram, Germany). PFR was measured by means of a quantum flat head PAR sensor (LI-190) connected to a radiometer (LiCor-250A Light Meter; Li-Cor Biosciences, Lincoln, USA). It corresponded to 65–70 μmol photons m⁻² s⁻¹ inside the water in the middle of the cylinder as measured with a spherical sensor (US-SQS/L, Walz, Germany). For each cylinder, about 1 g initial fresh weight (FW) of alga was used. Four cylinders containing each of them 4–5 independent specimens were used for each treatment. The physiological measurements described below were applied to fresh material taken directly from each cylinder at the end of the 7-day incubation period. Elemental composition was analysed from freeze-dried material stored at −80 °C.

Determination of the growth rate

Thallus growth was calculated by the difference between initial and final FW in each cylinder. Growth rate was estimated by fitting the exponential function, as proposed by Lüning (1985):

where r represents the intrinsic growth rate (% day⁻¹), P _t is the final fresh weight, P ₀ the initial fresh weight, and t the elapsed time in days.

Chlorophyll fluorescence

Optimal quantum yield for photosystem II (PSII) fluorescence (F _v/F _m) was measured by means of a pulse amplitude modulated fluorimeter using a Mini-PAM (Walz, Effeltrich, Germany) after 15 min of incubation in darkness, as described by Schreiber et al. (1986). Immediately afterwards, rapid light curves were measured using the same device, where the effective quantum yield of PSII (Φ_PSII = ΔF/F _m′) was estimated for 8 different white light irradiances provided by the internal halogen lamp. According to Genty et al. (1989), F _v is the maximal variable fluorescence of a dark-adapted sample, F _m the fluorescence intensity with all PSII reaction centres closed, F the fluorescence at any time during induction and F _m′ the light-saturated fluorescence. The electron transport rate between PSII and PSI (ETR) at each irradiance was calculated as:

where 0.5 stands for the assumption of equal contribution of excitons from PSI and PSII, and A is the thallus absorptance. The parameter A was estimated according to Beer et al. (2000), as the fraction of incident photons of photosynthetic active radiation (PAR) absorbed by the thallus:

where T is the transmittance and assuming no significant reflectance. The transmittance was determined by comparing readings from a quantum flat head PAR sensor (LI-190, Li-Cor Biosciences, Lincoln, USA) connected to a LiCor radiometer, with and without the thallus placed in the surface of the sensor, with a white led-lamp irradiating perpendicularly at a fixed distance.

ETR versus irradiance curves were fitted to the nonlinear least squares regression model by Eilers and Peeters (1988) using the Solver function of Excel (Microsoft, Redmond, USA) in order to obtain photosynthetic parameters: maximum electron transport rate (ETR_max), the initial slope of the curve related to the photosynthetic light-harvesting efficiency (α), the light requirement for saturating photosynthetic rate (E _k) which is given as the intercept between α and ETR_max, and the irradiance at which chronic photoinhibition begins (E _0pt).

All measurements were taken for each treatment (n = 4) after 7 days of cultivation using sterile FSW pre-equilibrated at either 390 or 1300 ppm CO₂.

Photosynthetic rates by O₂ evolution and ¹⁴C fixation

Pieces of blade of 50–100 mg FW used for measuring photosynthesis (by ¹⁴C fixation and oxygen evolution) and respiration were cut the day before measurements to avoid wound-healing interference. Sterile FSW pre-equilibrated at either 390 or 1300 ppm CO₂ was used for all measurements of photosynthesis/respiration.

Oxygen evolution

Net photosynthesis (NPS) under culture PFR provided by white light led-lamps, as well as dark respiration were estimated by oxygen evolution using a Clark-type oxygen electrode (5331; Yellow Spring Instruments, Ohio, USA) in 8-ml custom-made transparent Plexiglas chambers at 4 ± 0.2 °C. The water in the chambers was continuously stirred. Rate measurements were taken at 15-min intervals.

Inorganic ¹⁴C fixation

Measurements were based on the protocol proposed by Kremer and Küppers (1977). Samples were allowed to photosynthesise in a H¹⁴CO₃ ⁻ (Perkin Elmer, USA) medium or to assimilate ¹⁴C in darkness (as a control of light-independent carbon fixation). Thalli were placed in 8-ml septum-sealed glass vials filled with sterile FSW in a custom-made transparent Plexiglas container connected to a thermo-regulated water bath at 4 ± 0.2 °C, under culture PFR and with continuous homogenisation of the medium by a magnetic stirrer. A pre-acclimation period of 15 min in light for steady photosynthesis measurements and of 30 min in darkness for the dark control was applied before adding the H¹⁴CO₃ ⁻ solution. Subsequently, an aliquot of the H¹⁴CO₃ ⁻ stock was injected through the septum to yield a final specific activity of ~0.25 μCi μmol C⁻¹, and thalli were incubated during 30 min either in light or darkness. This incubation time was chosen in pilot experiments, trying to find the lowest time for the incubation which maintains enough sensitivity. After the incubation period, thalli were immediately rinsed in unlabelled medium, blotted, submersed in liquid nitrogen to stop reactions and settled in 20-ml scintillation vials containing 400 μl of HNO₃ (68 %); then, after 10 min at 50 °C, vials were left uncapped in an orbital shaker inside a fumehood until complete tissue solubilisation.

The equivalent carbon fixation rates were determined by measurements of acid stable ¹⁴C fixation after the addition of 10 ml of scintillation cocktail (Insta-gel, Perkin Elmer), using a liquid scintillation counter (TriCarb 2910, Perkin Elmer) with automatic quench correction, and referred to the total inorganic carbon content of the incubation media used. Activity levels of dark controls were subtracted from light ¹⁴C-fixation measurements. For the calculation of the carbon fixation rate, it was assumed that the uptake of ¹⁴C is 5 % slower than ¹²C (based on Steeman-Nielsen 1952).

Total carbon and nitrogen content

Total internal C and N contents were determined from freeze-dried tissue samples after homogenisation, using a C:H:N elemental auto-analyser (Perkin-Elmer 2400CHN) by DOI method (Kristensen and Andersen 1987).

Stable isotopic determination

The abundance of ¹³C relative to ¹²C in plant samples (c.a. 30 mg of dry mass) was determined by mass spectrometry using a DELTA V Advantage (Thermo Electron Corporation, USA) Isotope Ratio Mass Spectrometer (IRMS) connected to a Flash EA 1112 CNH analyser. The ¹³C isotopic discrimination in the algal samples (δ¹³C_alga) was expressed in the unit notation as deviations from the ¹³C/¹²C ratio of the Pee-Dee Belemnite CaCO₃ (PDB) calculated according to:

To determine isotopic composition of dissolved inorganic carbon (δ¹³C_DIC), 20 ml of FSW from each cylinder was filtered (Whatman GF/F), fixed and stored in septum-sealed glass vials without leaving a head-space. Measurements of δ¹³C_DIC were taken with the same IRMS connected to a GasBench II (Thermo Electron Corporation) system.

The δ¹³C_alga was corrected with the δ¹³C_DIC values from the medium, since the CO₂ source used in the experiment for the CO₂-enriched treatment came from previously fixed CO₂ which had been already discriminated.

DOC and POC

Samples for the determination of dissolved organic carbon (DOC) present in the growth medium were taken at the beginning and at the end of the incubation period and analysed by an automated system (TOC-L CSN, Shimadzu Corporation, Kyoto, Japan) according to the manufacturer’s protocols after filtration (Whatman GF/F). The filter was dried overnight at 80 °C and used for the determination of particulate organic carbon (POC) using the same elemental auto-analyser described above.

Data analyses

Replicate measurements (n = 4 independent thalli) were taken for significance of differences (P < 0.05) promoted by differences in CO₂ enrichment for each species using t tests. All statistical analyses were performed using the SigmaPlot 11.0 statistical software (Systat Software Inc., USA).

Table 1 shows the averaged values of the different variables of the seawater carbonate system in both CO₂ treatments, along the culture period. When pH was changed from 8.18 to 7.72 by aerating with CO₂-enriched air, dissolved inorganic carbon (DIC), dissolved CO₂ and HCO₃ ⁻ increased by 6, 204 and 8 %, respectively, and CO₃ ²⁻ decreased by 63 %, while TA showed no significant differences between low- and high CO₂ cultures.

Growth rate was significantly affected by the increase in dissolved CO₂ in both species, but not in the same direction. Rather, at high CO₂, growth rate of D. aculeata decreased by 82 % from 1.1 to 0.2 % day⁻¹ while in A. esculenta it increased by 34 % from 7.2 to 9.7 % day⁻¹ (Fig. 2). It must also be noted that A. esculenta exhibited a much higher growth rate than D. aculeata at both CO₂ treatments.

Growth rate expressed as % day⁻¹ (mean and SD, n = 4) of Desmarestia aculeata and Alaria esculenta during 7 days of culture at 390 and 1300 ppm CO₂. Statistically significant differences between both CO₂ conditions for each species are indicated by an asterisk (P < 0.05)

Full size image

In contrast to growth rate, gross oxygen production as well as C fixation rates did not show any significant differences between treatments for any of the two species (Fig. 3). Respiration rate decreased by 50 % from 10.6 to 5.3 μmol O₂ g FW⁻¹ h⁻¹ in A. esculenta at high CO₂ conditions, while it increased in D. aculeata from 3.1 to 4.9 μmol O₂ g FW⁻¹ h⁻¹. Desmarestia aculeata had a slight, but significantly lower net oxygen production rate at high CO₂, while A. esculenta exhibited no significant differences in net oxygen production rate. DOC release rate increased significantly in both species at high CO₂, especially in D. aculeata, with a value 5 times higher than at normal CO₂ conditions, from 0.1 to 0.49 μmol C g FW⁻¹ h⁻¹ in D. aculeata and from 0.23 to 0.57 μmol C g FW⁻¹ h⁻¹ in A. esculenta, while POC release rate did not show any significant differences between treatments (Fig. 4).

Net photosynthetic rate (a), gross photosynthetic rate (b) and dark respiration rate (d) measured by oxygen evolution and photosynthetic rate measured by ¹⁴C fixation (c) of Desmarestia aculeata and Alaria esculenta after 7 days of culture at 390 and 1300 ppm CO₂. Statistically significant differences between both CO₂ conditions for each species are indicated by asterisk (P < 0.05) (mean and SD, n = 4)

Full size image

Dissolved organic carbon (DOC) and particulate organic carbon (POC) release rate of Desmarestia aculeata and Alaria esculenta during 7 days of culture at 390 and 1300 ppm CO₂. Significant differences between both CO₂ conditions for each species are indicated by asterisk (P < 0.05) (mean and SD, n = 4)

Full size image

The relative contribution of the processes affecting the cellular carbon budget is shown in Fig. 5. Each process was recalculated to account for the organic C produced or consumed during the 7 days of cultivation for each cylinder (replicate) and is expressed as a percentage of the total amount of C fixed during that period (considering ¹⁴C fixation measurements as gross photosynthesis), which is invested in either respiration, new biomass production (growth), carbon accumulation in storage biomolecules not invested in growth, or DOC and POC released to the external medium. A constant light-dependent C fixation rate during the cultivation period and a constant respiration rate during light and dark periods over the 7 days of cultivation were assumed. The quantity of C accumulated in storage molecules was calculated as the difference in the total C-content of algal biomass (data from Table 4), corrected by the integrated DW, between the initial and the final time of the incubation period. In D. aculeata, the percentage of C fixed invested in growth strongly decreased at high CO₂ (from 25 to 5 % of total carbon fixed) at the expense of an increase in respiration, organic carbon accumulation and DOC release, whereas A. esculenta increased growth at high CO₂ (with an increment of around 15 % of total carbon fixed) as a result of a decrease in respiration and in organic carbon accumulation.

Distribution of total fixed carbon during the 7 days of culture at both CO₂ conditions among different cellular processes, expressed as a percentage. a fixed carbon lost by respiration; b invested in new biomass production (growth); c accumulated in storage biomolecules; d, e organic carbon released and found in the medium in dissolved (DOC) and particulate (POC) form, respectively

Full size image

Respect to the operation of photosynthesis, rapid light curves indicated a significant lower maximum electron transport rate (ETR_max) at elevated CO₂ conditions in D. aculeata (by 30 %), and a 35 % higher photosynthetic efficiency (α) and 60 % higher ETR_max at high CO₂ in A. esculenta (Table 2). Electron transport rate at 30 μmol photons m⁻² s⁻¹ (ETR₃₀), resembling culture conditions, showed significant differences between treatments in both species, with an increase in 45 % in A. esculenta and a decrease in 15 % in D. aculeata at increased CO₂, respect to normal CO₂ conditions. There was also a significant increase in the saturating irradiance (E _k) and the irradiance at which chronic photoinhibition begins (E _0pt) of about 25 % for A. esculenta at this condition, while D. aculeata did not show any significant differences in those parameters. Optimal quantum yield for PSII fluorescence (F _v/F _m) was not affected by CO₂ in any of the species, showing common optimal values for brown algae, thus indicating that thalli from both species and treatments were in an overall healthy state.

The photosynthetic quotient (PQ), calculated as the molar ratio of the rate of gross oxygen production to the rate of carbon dioxide fixation, was significantly lower at high CO₂ in A. esculenta, showing a decrease from 1.92 to 1.41, but it was not affected by CO₂ in D. aculeata, with values around 1.34 (Table 3).

The ¹³C isotopic discrimination in the algal samples (δ¹³C_alga), which was corrected with the isotopic composition of DIC in the medium, was significantly lower at high CO₂ in both species, with a decrease from −19.2 to −23.5 ‰ for D. aculeata and a decrease from −21.8 to −28.7 ‰ for A. esculenta (Table 4). FW:DW ratio significantly increased in A. esculenta at high CO₂, indicating a higher water content in this condition, while it did not change in D. aculeata. Elevated CO₂ levels affected the nitrogen metabolism of these species in opposite ways, decreasing the percentage of internal N in D. aculeata and increasing it in A. esculenta, whereas the percentage of internal C, as well as the C:N ratio, slightly decreased at high CO₂ only in A. esculenta (Table 4). Therefore, N-metabolism was enhanced at increased CO₂ conditions in A. esculenta, but not in D. aculeata.

The increase in dissolved CO₂ concentration has shown the potential to change the carbon yield and fate in common seaweed species of the Arctic in different ways. The negative effect of elevated CO₂ in the growth rate of D. aculeata has also been observed in other seaweeds such as the rhodophytes Hypnea musciformis (Israel and Hophy 2002), Porphyra leucosticta (Mercado et al. 1999) and Porphyra linearis (Israel et al. 1999), and the phaeophytes Fucus vesiculosus (Gutow et al. 2014) and S. latissima (Swanson and Fox 2007). As reason for this effect, the pH sensitivity of the specific CCMs of Porphyra species has been invoked (Moulin et al. 2011). For S. latissima, likewise, the authors proposed that the CCMs of this species allow for optimal photosynthesis at high seawater pH (Axelsson et al. 2000). Another possibility for the decrease in growth rate at high CO₂ conditions could be a lack of ability to equilibrate carbon fluxes with the metabolism of other nutrients such as N, as indicated by a lower total N-content, or a lack or malfunction of internal pH regulation. However, in our experiment, D. aculeata seemed not to be photochemically stressed as revealed by F _v/F _m values, unlike P. leucosticta (Mercado et al. 1999) and H. musciformis (Israel and Hophy 2002), which showed clear evidence of photochemical stress at increased CO₂ levels.

On the other hand, the positive effect of elevated CO₂ in the growth rate of A. esculenta has also been observed in many species of seaweeds such as the rhodophytes Lomentaria articulata (Kübler et al. 1999), Hypnea spinella (Suárez-Álvarez et al. 2012) and Neosiphonia harveyi (Olischläger and Wiencke 2013), the chlorophyte U. rigida (Gordillo et al. 2001) and the brown turf-forming alga Feldmannia sp. (Russell et al. 2009). This is the expected acclimation response of a photosynthetic organism to increased CO₂ when photosynthesis is not saturated at normal DIC level. Nevertheless, A. esculenta and D. aculeata did not change the C fixation rate under this condition (see Fig. 3c), indicating that their photosynthesis is already saturated at normal CO₂ levels, so the effect of CO₂ on growth does not correlate with the absence of effect on gross photosynthesis. The same pattern as in A. esculenta was obtained for U. rigida (Gordillo et al. 2001), in which the stimulation of growth was not caused by an increase in the photosynthetic rate; instead, the source of C for the extra biomass production came from the reduction in carbon losses.

DOC release has been proposed as a regulatory mechanism able to respond to the environment (Fogg 1983; Ormerod 1983), which would maintain the metabolic integrity of the cell and would protect the photosynthetic apparatus from an overload of products that cannot be used in growth or stored (Wood and Van Valen 1990). The release of DOC under high CO₂ levels increased in the unicellular green alga Dunaliella salina (Giordano et al. 1994) and in the coccolithophorid Emiliania huxleyi (Borchard and Engel 2012), while Hopkinson et al. (2010) and Sobrino et al. (2014) found that high CO₂ conditions reduced cellular carbon loss in natural phytoplanktonic communities. Furthermore, DOC release is suggested to act as a mechanism controlling the internal C:N ratio in the cyanobacterium S. platensis (Gordillo et al. 1999), in U. rigida (Gordillo et al. 2001) and in coastal phytoplankton (Sobrino et al. 2014). In this way, both species of the present study seem to increase the release of DOC at high CO₂ conditions, although the percentage of extracellular release (PER) was always in the range between 0 and 10 % of photosynthetic carbon assimilation (see Fig. 5, d), as it has been shown for healthy growing algae (Sharp 1977; Mague et al. 1980). In D. aculeata, the decrease in total N-content at high CO₂ levels, presumably due to a decrease in the assimilation rate of N, could lead to an enhancement in DOC release, thus maintaining the internal C:N ratio. However, in A. esculenta, the C:N ratio is not maintained between treatments, due to an increase in DOC release at enriched-CO₂ levels together with an enhancement of the uptake and assimilation of N, as is suggested by the higher N-content values.

Changes in dark respiration rates are in accordance with the growth rate response. Previous evidence of the decrease in respiration rate by high CO₂ concentrations has been reported in vascular plants (Bunce and Caulfield 1991; Azcón-Bieto et al. 1994) and in U. rigida (Gordillo et al. 2001), although there were no changes in respiration rates in the rhodophytes P. leucosticta (Mercado et al. 1999), H. spinella (Suárez-Álvarez et al. 2012), Gracilaria lemaneiformis (Zou and Gao 2009) and Hizikia fusiformis (Zou et al. 2011a). On the contrary, an increase in the respiration rate was recorded in some vascular plants (Davey et al. 2004) and in the diatom Thalassiosira pseudonana (Yang and Gao 2012). Stimulation of mitochondrial activity under OA scenarios may be associated with altered proton gradients across the mitochondrial membrane or to pH-dependent changes in the functioning of respiratory enzymes (Amthor 1991). On the other hand, down-regulation of CCMs would leave the cell with excess energy equivalents and so, cells could down-regulate energy production by reducing mitochondrial respiration (Hennon et al. 2014).

Regarding the regulation of the photochemical performance, the significant differences obtained in ETR_max between CO₂ treatments in both species could be the result of a change in the number of active reaction centres (see Table 2). Furthermore, the differences obtained in ETR₃₀, in contrast with the absence of effect on gross oxygen production and C fixation in both species, could be due to the alteration in light capture efficiency by PSI and PSII at high CO₂ conditions, as reported by Satoh et al. (2002), along with a change in the photosynthetic pigment content and/or antennae size. In A. esculenta, an activation of the cyclic electron flow around PSII as part of a photoprotection strategy at saturating irradiances could explain the increase in ETR₃₀ at high CO₂ levels, as suggested for U. rigida by Gordillo et al. (2003) and for P. tricornutum by Feikema et al. (2006). In this way, an enhancement of photoprotection mechanisms at high CO₂ in A. esculenta could be responsible for the significant increase in E _k and E _0pt (Table 2). In D. aculeata, increased acidity in the ambient seawater might, to some extent, affect intracellular acid–base balance, and hence, cause a decrease in ETR. The same response has been observed in the haptophyte Phaeocystis globosa at the beginning of the acclimation period to elevated CO₂ (Chen et al. 2014), although the photosynthetic efficiency was not affected. It is also necessary to take into account that the calculation of the gross oxygen evolution, which was estimated assuming a constant respiration rate as measured in darkness, could lead to an overestimation since respiration may decrease in light, as it has been previously shown for some seaweeds (Brown and Tregunna 1967, Zou et al. 2011b).

The C fixation rate was not affected by high CO₂ in any of the two species studied, indicating that photosynthesis is carbon saturated at current CO₂ concentrations due to CCM functioning, as it is suggested by δ¹³C isotopic discrimination data. Algae with δ¹³C values more negative than −30 ‰ are unable to increase the pH of seawater above 9.0 (Maberly et al. 1992), which indicates the absence of functional CCMs (photosynthesis relies only on diffusive-CO₂ entry). For A. esculenta, Maberly et al. (1992) published a δ¹³C value of −17.8 ‰, while for D. aculeata δ¹³,C values are between −18 and −26 ‰ (Raven et al. 2002), being these values similar to the ones obtained in the present study at normal CO₂ levels for both species (Table 4).

A significant increase in the carbon isotope discrimination data, i.e. more negative δ¹³C values relative to the PDB standard, indicates a deactivation of CCMs in both species at high CO₂ conditions, as reported for other algae (Hurd et al. 2009; Cornwall et al. 2012). In A. esculenta, the energy saved due to CCM deactivation could be invested in other processes like assimilation of other nutrients or increasing synthesis of biomolecules, giving an increase in the growth rate, as it has been described for U. rigida (Gordillo et al. 2001). In D. aculeata, CCM deactivation did not produce an increase in growth rate as ETR₃₀ decrease, probably due to the sensitivity of this species to a lower pH. However, δ¹³C data should be taken cautiously, since a higher ¹³C-discrimination not only indicates a higher use of diffusive-CO₂ but also could indicate a higher photosynthetic dependence on the CO₂ supplied from HCO₃ ⁻ via external carbonic anhydrase, as this enzyme could also be subjected to kinetic fractionation (Mercado et al. 2009; O’Leary et al. 1992).

Moreover, PQ (which reflects de efficiency of the photochemical reactions from the water reduction occurring in PSII to C fixation) was significantly lower in A. esculenta at high CO₂ conditions (Table 3), indicating a lower amount of photosynthetic energy invested in processes other than C fixation, although energy obtained by cyclic electron flow around PSII, which is suggested as a photoprotective mechanism at high CO₂ in this species, is not taken into account by O₂ production. These results are in accordance with the deactivation of CCMs indicated by ¹³C isotopic discrimination data, although D. aculeata did not show significant differences in PQ values, suggesting that the energy saved due to CCM deactivation may be invested in counteracting the external pH reduction (Wu et al. 2010).

The carbon balance of D. aculeata showed the typical pattern of physiological performance of brown seaweeds in the Arctic during summer. This might explain why the growth is low (25 % of total C fixed) and the organic carbon accumulation is high (64 %) at control conditions. It has been reported in some species that this pattern is under the control of photoperiodisms or endogenous free-running circannual rhythms entrained by a critical minimum daylength in autumn (Lüning 1991; Schaffelke and Lüning 1994; Wiencke et al. 2011), suggesting that the addition of nitrate to summer N-limited brown seaweeds would have only a marginal effect on growth and biochemical composition (Henley and Dunton 1997). Nevertheless, A. esculenta responded in a different way, with a high investment of carbon into growth (72 % of total C fixed) and a low organic carbon accumulation (14 % of total C fixed) at normal CO₂ levels. It is also possible that the age of the thalli used for the experiment may have influenced this different pattern, together with the morphological differences between both species. Sporophytes of D. aculeata have a trichothallic growth in which cell division is restricted to the base of one or several filaments, and the apex of a branch ends with a single filament having an intercalary meristem which produce lateral filaments above and below it (Bold and Wynne 1978), while sporophytes of A. esculenta have an intercalary growth with a meristem situated at the juncture of the stipe and the blade. However, it has been shown in other studies that both species could reach similar growth rates (Bischoff and Wiencke 1993). Also, young thalli could have a metabolism more focused on growing than mature thalli, although a more plausible explanation for the different growth rates would be that growth in A. esculenta is not controlled by this endogenous circannual rhythm or photoperiodism, as demonstrated by Gordillo et al. (2006). The previous study showed a significant change in the nitrate reductase (NR) activity and in the internal N content at N-enriched conditions compared to N-depleted conditions in the Arctic summer for A. esculenta, while D. aculeata showed no response in those variables.

The decrease in total N content obtained in D. aculeata indicates a lower investment of energy in the N metabolism, as shown for Gracilaria tenuistipitata (García-Sánchez et al. 1994) and G. lemaneiformis (Xu et al. 2010), while A. esculenta exhibited an increase in total N content, suggesting a stimulation of N assimilation, which could be due to a higher synthesis of aminoacids necessary for a higher growth rate. Stimulation of N uptake and assimilation (via increasing NR activity) at high CO₂ conditions has been previously reported in the seaweeds P. leucosticta (Mercado et al. 1999), U. rigida (Gordillo et al. 2001), Chondrus crispus and Cystoseira tamariscifolia (Olabarria et al. 2012), and in H. spinella (Suárez-Álvarez et al. 2012). Moreover, NR activity was significantly enhanced in A. esculenta at increased CO₂ levels in a similar experiment carried out also in Kongsfjorden (Gordillo et al. unpublished results), which support our results of an increase in total N content, while NR activity from D. aculeata did not change between treatments.

Overall, the effect of CO₂ involves a reorganization of the energetic and carbon budget of the cell and does not reflect a direct effect on photosynthesis in the analysed species, but rather a significant effect on respiration, organic carbon accumulation and DOC release (see Fig. 5), along with an effect on CCMs and N metabolism, which ultimately determine growth rate. It has been reported that phytoplankton species apparently possess sensory systems that respond to environmental CO₂ concentrations and control the CO₂ acquisition efficiency (Matsuda et al. 2001; Burkhardt et al. 2001), suggesting that CO₂ plays more roles than simply being a substrate for photosynthesis.

The sum of the percentages of all processes taken into account in the carbon balance gave a total value higher that 100 % for both treatments and species, indicating a possible underestimation of the total C fixed measured by the ¹⁴C-method; this could be due to the respiration of some of the ¹⁴C-molecules which came from ¹⁴CO₂ already fixed by cells during the 30-min incubation period, but could also be due to the assumption of a constant respiration rate as measured in darkness, which could lead to an overestimation for the light period as mentioned above.

In the CO₂-enriched global scenario predicted, it can be concluded that seaweeds are going to be affected in different species-specific ways. Alaria esculenta might be benefitted and D. aculeata seems to be a clear loser. As a result, the Arctic seaweed community may shift its relative biomass dominance and the consequences propagate to the rest of the trophic web. Moreover, it is necessary to take into account that interactive effects of CO₂ and other variables such as nutrients, temperature, salinity and irradiance levels may influence their response and, thus, complicate the prediction of implications of OA for seaweeds.

Recently, there is a wide discussion regarding the duration of the exposure to increased CO₂ conditions in laboratory experiments. Ten days of incubation has been shown to be enough time to see a physiological acclimation of different species of seaweeds in previous studies, although experiments of longer-term acclimation are also needed to extrapolate their responses to the future environment, as there could be complex effects different from short-term physiological responses. Also, the responses to a more realistic gradual increase in CO₂ are largely unknown (but see Hall-Spencer et al. 2008). However, highly controlled short-term single or multifactorial laboratory experiments, like the present study, are also important in identifying the species’ preadapted sensitivities to increasing CO₂ (Roleda and Hurd 2012).