A novel in situ system to evaluate the effect of high CO 2 on photosynthesis and biochemistry of seaweeds

Previous studies of the impact of increased CO2 on macroalgae have mainly been done in laboratories or mesocosm systems, placing organisms under both artificial light and seawater conditions. In this study, macroalgae were incubated in situ in UV-transparent cylinders under conditions similar to the external environment. This system was tested in a short-term study (5.5 h incubation) on the effect of 2 partial pressures of CO2 (pCO2): air (ambient CO2) and the pCO2 predicted by the end of the 21st century (700 μatm, high CO2), on photosynthesis, photosynthetic pigments and photoprotection in calcifying (Ellisolandia elongata and Padina pavonica) and non-calcifying (Cystoseira tamariscifolia) macroalgae. The calcifying P. pavonica showed higher net photosynthesis under high CO2 than under ambient CO2 conditions, whereas the opposite occurred in C. tamariscifolia. Both brown algae (P. pavonica and C. tamariscifolia) showed activation of non-photochemical quenching mechanisms under high CO2 conditions. However, in P. pavonica the phenol content was reduced after CO2 enrichment. In contrast to phenols, in E. elongata other photoprotectors such as zeaxanthin and palythine (mycosporine-like amino acid) tended to increase in the high CO2 treatment. The different responses of these species to elevated pCO2 may be due to anatomical and physiological differences and could represent a shift in their relative dominance as key species in the face of ocean acidification (OA). More in situ studies could be carried out to evaluate how macroalgae will respond to increases in pCO2 in a future OA scenario. The in situ incubator system proposed in this work may contribute towards increasing this knowledge.


INTRODUCTION
The ocean absorbs about 30% of the emitted anthropogenic CO 2 , causing significant changes in the marine carbon cycle and carbonate system.These changes include an increase in the concentration of dissolved CO 2 , a smaller proportional increase of bicarbonate ions (HCO 3 ions (CO 3 2− ) and changes in the saturation state of calcium carbonate (CaCO 3 ) (Caldeira & Wickett 2005, Orr et al. 2005).The predicted changes in dissolved inorganic carbon distri bution and abundance will result in an increase in hydrogen ion (H + ) concentration and, consequently, a decrease in seawater pH.These interrelated chemical changes in the inorganic carbon system are referred as 'ocean acidification (OA)' (Zeebe et al. 2008, Shi et al. 2009).The pH of ocean surface waters has decreased by 0.1 units since the beginning of the industrial era to its current mean value of 8.2 (Caldeira & Wickett 2003), corresponding to a 26% increase in H + concentration.The most recent models project a global increase in OA, with a corresponding decrease in surface ocean pH by the end of 21st century in the range of 0.06−0.07, up to 0.30−0.32,depending on the case scenarios of CO 2 atmospheric concentration (Representative Concentration Pathways, RCP) (IPCC 2013).
While the chemistry of carbonate systems has been well studied (Zeebe 2012), the impacts of OA on marine organisms and ecosystems remain poorly understood (Gattuso et al. 2010).While many studies have pointed out that OA may have negative effects on macroalgae (Mercado et al. 1999, Riebesell et al. 2000, Zondervan et al. 2001, Hall-Spencer et al. 2008, Kuffner et al. 2008, Gao & Zheng 2010, Sinutok et al. 2011, Johnson et al. 2012), a few found no effects (Israel & Hophy 2002, Egilsdottir et al. 2013) and others even found in creases in photosynthesis, growth and calcification rates (Gao et al. 1993, Kübler et al. 1999, Zou 2005, Iglesias-Rodríguez et al. 2008, Zou & Gao 2009, Roleda et al. 2012).Many factors may be involved in the discrepancies among results, such as morpho-functional traits, the form of C acquisition, species adaptation/ acclimation to environmental con ditions, pre-experimental conditions and other methodological aspects (Ries et al. 2009, Olabarria et al. 2013).In addition, most of these studies have been carried out under laboratory conditions and in different temporal scales (see Hurd et al. 2009 and Raven 2011 for a review), impeding an appropriate comparison among the results.Although mesocosm experiments have improved our understanding of how submerged macro phytes will respond to future OA, information from in situ experiments is scarce.To date, in situ responses of marine organism to dissolved CO 2 increase are based on observations made near volcanic vents (Hall-Spencer et al. 2008, Martin et al. 2008, Porzio et al. 2011, Johnson et al. 2012).Recently, different techniques of in situ CO 2 manipulation, such as the Coral−Proto Free Ocean Carbon Enrichment System (Kline et al. 2012), the Free Ocean Carbon Enrichment System (Arnold et al. 2012), and the Carbon-Enriched Open Chamber System (Campbell & Fourqurean 2011, 2013) have been proposed.The development of new CO 2 in situ experiments can provide new insights and give straightforward answers or at least provide a piece of the puzzle about the effect of OA on macroalgae.
The objective of this study was to present a novel and simple experimental design to incubate macroalgae in situ under different partial pressures of CO 2 (pCO 2 ).Our design was tested in the lower intertidal environment of Cabo de Gata National Park (Spain), by comparing the responses of non-calcifying (Cystoseira tamariscifolia) vs. calcifying (Padina pavonica and Ellisolandia elongata) macroalgal species to changes in pCO 2 .The studied calcifying species are lightly calcified with aragonite (P.pavonica) and heavily calcified with magnesium calcite (E.elongata).We analyzed the short-term effects of in creased pCO 2 in photosynthetic parameters (both O 2 evolution-and chlorophyll fluorescence-based para meters), as well as pigment and photoprotector concentrations (mycosporine-like aminoacids [MAAs] and phenolic compounds); all of which are good indicators of physiological status (Figueroa & Korbee 2010).
The brown macroalgae C. tamariscifolia is an Atlantic species that occurs from Scotland and Ireland to Mauritania and Cape Verde Islands.The species is present in Mediterranean waters of Atlantic influence, occurring across Iberian Coast as far as the province of Almería (Gómez-Garreta et al. 1994).It presents blue-green iridescence, can reach up to 1 m height and is fixed to the substrate by a thick disk (Gómez-Garreta et al. 2001).The brown macroalgae P. pavonica is a highly spread tropical/subtropical species, common in the Mediterranean coastal waters (www.algaebase.org).The thallus is brown to tan in colour, forming fan-shaped clusters.The blades are calcifying, heavier above and lighter below, and curl inward near the edges.Both the upper and lower blade surfaces bear minute surface hairs arranged in a series of bands approximately 1.5 to 6 mm apart (Taylor 1979, Littler & Littler 2000, Littler et al. 2008).The blades attach to the substratum via a holdfast, which is often matted.The red E. elongata is an articulated calcareous species, whitish-pink to reddishlilac, calcified, up to 50 mm high.The species is present in Medi ter ranean and Eastern Atlantic waters (www.algaebase.org).

Incubation and experimental design
The studied macroalgae were collected at 0.5 m depth and immediately incubated in 6 transparent UV cy linders (0.8 l; 33 × 7 cm) in a sheltered bay.The cylinders were fixed perpendicularly to the sun and not deeper than 0.5 m using a wooden frame and several buoys and weights (Fig. 1).Two pCO 2 conditions were applied in triplicate cylinders: (1) bubbling air (control, ambient CO 2 ) and (2) a commercial mix (Praxair España) of CO 2 and air with a final concentration of 700 ppm (high CO 2 ).The air pump and the bottle with 700 ppm CO 2 were maintained on the beach edge.Aeration was provided from the bottom of the cylinders at a rate of 0.5 l min −1 .Before algal incubation, the water was aerated for 30 min in both pCO 2 treatments.In a previous experiment it was determined that this time was enough to equilibrate the carbonate system in side the cylinders.Each cylinder received 20 g fresh weight (FW) of algae.The incubation experiments were performed over 3 d in September 2012, with 1 d for each species.
After bubbling for 30 min, aeration was stopped in the 6 experimental cylinders in order to avoid extra oxygenation, and 1 h incubation allowed for net photosynthesis determination.This time is optimal for the incubations in function of the high volume: biomass incubation ratio inside the cylinders.At the end of this time, water samples were collected for the final oxygen concentration.After that, the cylinders were closed and were continuously aerated for 5.5 h.
Photosynthetic parameters including maximal quantum yield (F v /F m ), electron transport rate (ETR) and non-photochemical quenching (NPQ) were measured by in vivo chlorophyll a (chl a) fluorescence associated to Photosystem II at the beginning and at the end of the incubation period of 5.5 h.Samples for absorptance, chlorophylls, carotenoids, photo protective compounds (phenols and myco spo rine-like amino acids) as well as antioxidant capacity were also determined at both periods.
To analyse phenolic compounds, DPPH and photosynthetic pigments, samples were collected, immediately frozen in liquid nitrogen and stored at −80°C until analyses.Samples for MAAs were kept desiccated until analysis.Water samples for pH measurement and alkalinity analysis were obtained at the end of the 5.5 h of incubation; the samples for alkalinity were poisoned with a small amount of saturated mercury chloride solution until analysis.The water chemistry for incubation without algae was also determined.

Measurement of solar radiation, temperature and nutrients
The irradiance of solar radiation was determined at 3 wavelength bands (PAR = 400 to 700 nm, UVA = 315 to 400 nm and UVB = 280 to 315 nm) using 2 Hyperspectral Irradiance Sensors for UV and PAR (Ramses, TrioS).Due to the sensor size, irradiances were measured outside the cylinders.
Temperature was measured during the incubation periods inside the cylinders.The concentration of nitrate and phosphate was measured in the seawater, and after the incubation it was also determined inside of the cylinders.

Measurement of pH and salinity
The pH within the incubation cylinders was measured using a pH meter (Crison Basic 20, Crison Instruments).The pH electrode was calibrated regu-Fig.1. Experimental design: cylinders were fixed using a wooden frame larly with standard National Bureau of Standards (NBS) buffer solutions (Oakton) to ensure a stable response.Salinity was estimated using a conductivity meter (Crison CM35, Crison Instruments).

Total alkalinity
Total alkalinity was measured by titrating (stepwise addition of reagent) the water sample with HCl to a final pH NBS of 3. Once the water sample reached a pH of 3, all the bicarbonate, carbonate and hydroxide were neutralized.An automated titration system (877 Titrino plus, Metrohm) was used selecting the option of monotonic titrations with automatic equivalence point finding (MET).
The HCl solution nominal molarity used was 0.97 M at 20°C.This molarity was verified titrating 80 ml of NaHCO 3 .The total inorganic carbon was determined using continuous and in situ measurements or pH and temperature and determina tions of total alkalinity and salinity.The total alkalinity was determined by the Gran (1952) titration method.Total inorganic carbon concentration was calculated using the program CO 2 sys (v.2.1, Pierrot et al. 2006) using dissociation constants for car bonic and boric acids determined on the NBS scale.To calculate the speciation of total inorganic carbon into carbonate, bicarbonate and dissolved CO 2 forms, the CO 2 seawater solubility coefficient proposed by Weiss (1974) was used.The first and second dissociation constants of carbonic acid in sea water by Mehrbach et al. (1973), refit by Dick son & Millero (1987), and the first dissociation constant of boric acid in seawater by Lyman (1956) were used.

Photosynthetic measurements as oxygen evolution
Net photosynthesis was estimated as photosynthetic oxygen evolution under in situ incubation by the difference between the oxygen concentrations after (final) and before (initial) 1 h incubation under the 2 pCO 2 treatments.The Spectrophotometric Winkler method was used to estimate the concentration of dissolved oxygen (Labasque et al. 2004).In each case, after fixing the soluble oxygen with R1 and R2 Winkler reagents, samples were kept in darkness and at 4°C.Within 24 h of collection, R3 was added and absorbance was measured at 466 nm, using a Genesis 10S Vis Thermo Scientific (Thermo Fisher Scientific).Standardization relied on the preparation of I2+I3 solutions by oxidation of iodide with iodate.A standard solution KIO 3 (0.01M) was used to obtain the standard curve.

In vivo chl a fluorescence
In vivo chl a fluorescence associated to Photosystem II was determined by using a portable pulse modulated fluorometer (Diving-PAM, Walz).Algal samples were collected from each treatment at the initial time, and after 5.5 h incubation were put into 10 ml incubation chambers to conduct rapid light curves (RLCs).For these incubations, the medium was taken directly from each of the cylinders.Minimum (F o ), maximum (F m ) and maximum variable fluorescence (F v = F m − F o ) were determined after 15 min in darkness to obtain F v /F m (Schreiber et al. 1995, Figueroa et al. 2003).
The ETR (µmol electrons m −2 s −1 ) was calculated according to Schreiber et al. (1995) as follows: where E is the incident irradiance.ΔF = F m ' − F t and is the variable fluorescence in light, F m ' is the maximum fluorescence in light and F t is the intrinsic fluorescence under a specific irradiance.Absorptance, Maximal NPQ (NPQ max ) and the initial slope of NPQ versus light curve (α NPQ ) were obtained from the tangential function according to Eilers & Peeters (1988).Finally, the saturation light for NPQ (Ek NPQ ) was calculated from the intercept between NPQ max and α NPQ .

Photosynthetic pigments
Photosynthetic pigments were measured for each species and cylinder in duplicate.Results were expressed as mg g −1 DW (dry weight).The FW:DW ratios were calculated from 10 thalli of each species.FW was determined after blotting off surface water with absorbent paper.Afterwards, the thalli were oven-dried for 2 d at 60°C to obtain the DW.The ratios were 4.3 ± 0.08 for C. tamariscifolia, 3.1 ± 0.04 for P. pavonica and 1.5 ± 0.01 for E. elongata.
Chl a content was determined spectrophotometrically (Shimadzu UVmini 1240, Shimadzu Scientific Instruments), while chl c and carotenoids were identified and quantified by high-performance liquid chromatography (HPLC, Waters 600 HPLC system, Waters Cromatografía).Both analyses were made from the same extract using 15 mg FW in 1 ml of N,Ndimethylformamide (DMF) and maintained in darkness at 4°C for 12 h.The chl a concentration was calculated using Wellburn (1994) equations.The carotenoid composition and concentration were determined by HPLC according to García-Sánchez et al. (2012).Chl c, fucoxanthin, violaxanthin, antheraxanthin, zeaxanthin and β-carotene were identified using commercial standards (DHI LAB Products).

Phenolic compounds
The phenol concentration was determined using 0.25 g FW.Samples were pulverized in a mortar and pestle with sea-sand using 2.5 ml of 80% methanol.After being maintained overnight, the mixture was centrifuged at 2253 × g for 15 min at 4°C and the supernatant was collected.These supernatants were used for phenol determination and to determine the anti oxidant activity.Total phenolic compounds were estimated colourimetrically using Folin-Ciocalteu assay (Folin & Ciocalteu 1927).Phloro glucinol (1, 3, 5-trihydroxybenzene, Sigma P-3502) was used as standard.Finally, the absorbance of 760 nm was determined in a UVmini-1240 spectro photometer (Shimadzu Scientific Instruments) (Abdala-Díaz et al. 2006).

Antioxidant activity
Antioxidant activity of the algal extracts of C. tamariscifolia and P. pavonica was determined by DPPH (2, 2-diphenyl-1-picrylhydrazyl) free radical assay (Blois 1958).A volume of 150 µl of DPPH solution (diluted in 90% methanol) was added to each algal extract, obtained as previously explained above for phenols.The samples were reacted with the stable DPPH solution during 30 min in the dark at room temperature (~20°C).The absorbance of the solutions was read at 517 nm in a UVmini-1240 spectrophotometer.A calibration curve made for a set of DPPH con centrations was used to calculate the remaining con centration of DPPH in the reaction mixture after incubation.Concentrations of DPPH (mM) were plotted against plant extract concentration (mg ml −1 DW) in order to obtain the EC 50 value (oxidation index), which represents the concentration of the extract (mg ml −1 ) required to scavenge 50% of the DPPH in the reaction mixture.Ascorbic acid was used as positive control (Connan et al. 2006).

MAAs
Samples (10 to 20 mg DW) of E. elongata were extracted for 2 h in screw-capped centrifuge vials filled with 1 ml 20% aqueous methanol (v/v) at 45°C.The concentration and composition of different MAAs were analysed by HPLC (Waters 600 HPLC system, Waters Cromatografía) according to Korbee-Peinado et al. (2004).

Statistical analysis
The effects of the treatments (ambient CO 2 and high CO 2 ) on the photosynthetic parameters were ana lysed using ANOVAs (α = 0.05) (Underwood 1997).One test was performed including CO 2 treatment as fixed factor with 2 levels, and cylinder as a random factor nested within CO 2 treatment.Homogeneity of variance was tested using Cochran's tests and by visual inspection of the residuals (Underwood 1997).Photosynthetic parameters, phenolic compounds, antioxidant capacity and photosynthetic pigments were measured from 2 replicates in each cylinder (n = 12).For oxygen evolution, 4 measurements were done in each cylinder (n = 24); for the other variables (alkalinity, pH, temperature and carbonate chemistry) 1 measurement was done per cylinder (n = 6).Therefore, this source of variation (i.e.cylinder), is not included in the latest analysis.Analyses were done with SPSS v.21 (IBM).All results are expressed as mean ± SE.

Physical and chemical variables
Irradiance and temperature within the cylinders showed little variation during each species' incubation period, but changed along the 3 experimen tal days on which each species was incubated (Table 1).Daily integrated irradiance was higher during the Cystoseira tamariscifolia experiment, com pared to Padina pavonica and Ellisolandia elongata.Water temperature was lower during the E. elongata experiment than on the other 2 experimental days (Table 1).
After 30 min of pumping high (700 ppm) and ambient CO 2 inside the cylinders without algae, values of DIC were significantly higher in high CO 2 (2.2 ± 0.01 mM) compared to ambient CO 2 (2.0 ± 0.003 mM) (F = 135.42,df = 1, n = 6, p = 0.00031).Values of pH were lower in high CO 2 (8.0 ± 0.01) compared to ambient CO 2 (8.3 ± 0.01) (F = 539.35,df = 1, n = 6, p = 0.00002), while the total alkalinity did not show differences between treatments (2424 ± 10 and 2450 ± 10 µM, for high and ambient CO 2 , respectively; F = 4.49, df = 1, n = 6, p = 0.101).After algae incubation, the carbonate chemistry changed with respect to the above-reported values (Table 1).No differences in pH were found anymore between ambient and high CO 2 treatments in E. elongata and P. pavonica whereas the difference in pH in C. tamariscifolia was still evident after incubation (Tables 1 & 2).The pH increased with respect to the value found without algae, except for E. elongata at ambient CO 2 (Table 1).
During incubation, nutrient concentrations in the seawater were 1.4 µM nitrate and 0.1 µM phosphate.After incubation, the concentration of phosphate slightly decreased inside the cylinders, while nitrate decreased to values in the range of 0.2 to 0.4 µM in all species and both pCO 2 treatments.

Photosynthetic parameters
The results of net photosynthesis (determined by oxygen evolution) showed that, under ambient CO 2 conditions, net productivity rates of C. tamariscifolia (0.54 ± 0.02 mg O 2 g −1 DW h −1 ) were twice as high as values found for P. pavonica (0.23 ± 0.01 mg O 2 g −1 DW h −1 ), and 10 times higher than values for E. elongata (0.02 ± 0.003 mg O 2 g −1 DW h −1 ).When the different species were incubated in high CO 2 conditions, C. tamariscifolia showed a significant reduction in net photosynthesis, while P. pavonica showed a significant increase in this parameter (Fig. 2, Table 3).E. elongata showed a slight increase in photosynthetic rates under high CO 2 , but differences between the treatments were not significant (Fig. 2, Table 3).No significant differences were found between field samples (initial values) and those for the control treatment (ambient CO 2 ) in any of the photosynthetic or biochemical variables, so initial values are not shown.In the ambient CO 2 treatment, values of F v /F m registered for C. tamariscifolia, P. pavonica and E. elongata were 0.64 ± 0.02, 0.61 ± 0.03 and 0.58 ± 0.02, respectively (Table 4).The analysis of the RLCs showed that, under ambient CO 2 conditions, C. tamariscifolia and P. pavonica had higher values of ETR max , α ETR and Ek ETR compared to the calcareous red alga E. elongata.Values of ETR max for C. tamariscifolia and P. pavonica were 18 and 16 times higher than for E. elongata.The latter also showed a fall in photosynthetic rates under irradiances above 500 µmol m −2 s −1 (Fig. 3).NPQ parameters were also higher for the brown algae than for the red calcareous alga.An exception was found for α NPQ , where in E. elongata it showed higher values than that in the other 2 species.When the seaweeds were incubated in high CO 2 , increases of F v /F m and NPQ max in C. tamariscifolia and of α NPQ in P. pavonica were observed compared to the control (Fig. 3, Table 4).Nevertheless, the tendency towards higher values for NPQ max and/or α NPQ under high CO 2 was found for all 3 species.

Biochemical analysis
The analysis of photosynthetic and photoprotective pigments showed that C. tamariscifolia presented the highest concentrations of analysed pigments (chlorophylls, fucoxanthin, violaxanthin, antheraxanthin and β-carotene) (Table 5).Among the brown algae, concentrations of chl a, chl c and fucoxanthin were about twice as high in C. tamariscifolia compared to P. pavonica.The red alga E. elongata showed the lowest concentration of chlorophylls and carotenoids among the 3 species, particularly for fucoxanthin and violaxanthin (Table 5).In addition, a higher proportion of antheraxanthin+zeaxanthin to violaxanthin con tent were found in E. elongata compared to the 2 brown algae (Table 5).After the incubation period, the concentration of some photosynthetic pigments (chl a, chl c and/or phycobiliproteins) did not change significantly between ambient and high CO 2 treatments (Table 3).A significant difference between treat ments was found for E. elongata, in which the zea xanthin content was significantly higher under high CO 2 than under ambient CO 2 conditions (Tables 3 & 5).
Among the brown algae, C. tamariscifolia showed higher phenol content (25 mg g −1 DW) and antioxidant capacity than P. pavonica (20 mg g −1 DW) (Fig. 4).Significant differences in phenols and antioxidant capacity between treatments were only found for P. pavonica, in which a decrease in phenolic compounds was observed under high CO 2 (Fig. 4, Table 3).A negative correlation was found between 251 Fig. 2. Net photosynthesis in Cystoseira tamariscifolia, Padina pavonica and Ellisolandia elongata after 1.5 h in situ incubation in high CO 2 and ambient CO 2 .Data are expressed as mean values ± SE (n = 24).Lowercase letters denote significant differences (ANOVA, p < 0.05) The content of total MAAs found in the red alga E. elongata was not sig nificantly different between either CO 2 treatment (0.62 ± 0.06 mg g -1 DW for high CO 2 and 0.66 ± 0.07 mg g -1 DW for ambient CO 2 , n = 12) (Table 3).The MAA composition was also similar between treatments as follows: 50 to 60% shinorine, 40% palythine and 5 to 10% asterina-330 (Fig. 6, Table 3).

DISCUSSION
Most studies on the effects of OA on marine macrophytes have been conducted ex situ in laboratories or mesocosms under controlled conditions.While those studies are useful to better understand some isolated effects of increasing dissolved CO 2 on algal photosynthesis and biochemistry, the results do not reflect the response of the natural populations.In situ experimental approaches can operate under more realistic  The results showed that the short-term responses of the seaweeds to seawater CO 2 enrichment varied according to species and their functional traits.While the brown algae C. tamariscifolia showed a reduction in photosynthetic rate (based on O 2 evolution), the calcified brown algae P. pavonica showed an increase under high CO 2 conditions.Both algae enhanced the NPQ mechanisms, but no changes in pigment composition or concentration were found.The enhanced production by P. pavonica under high CO 2 came with reductions in phenol.The calcareous red algae E. elongata was not significantly affected by CO 2 enrichment in most of the photosynthetic and biochemical parameters, but it showed increases in the concentration of its photoprotective pigment, zea xanthin.The contrasting results found for the 3 studied species in response to pCO 2 enrichment may be related to the striking differences in the photosynthetic apparatus, including pigment composition and concentration, and their highly distinct morphological traits and growth strategies.Among the brown algae, C. tamariscifolia has thick blades, with highly corticated and complex thallus, compared to the simpler sheet-like, thinner P. pavonica.The articulated calcareous E. elongata has the lowest ratio of photosynthetic to non-photosynthetic (calcified) tissue.
At the end of the incubation period, the effect of pCO 2 on seawater carbonate chemistry and pH was different depending on the incubated species.These may be related to differences in metabolism, since seawater carbonate chemistry is strongly affected by biological activity (Feely et al. 2004, Raven 2011)  While calcareous algae are known to be vulnerable to OA, by decreasing their calcium carbonate fixation and increasing dissolution (Feely et al. 2004), the non-calcified species are likely to show positive responses to elevated pCO 2 , since many macroalgal species have been shown to be carbon-limited in nature (Mercado et al. 1999, Koch et al. 2013), in part due to the inefficacies of Rubisco and inorganic carbon uptake mechanisms, as well as boundary layer-related problems (Raven et al. 2012).However, the degree of macroalgae photosynthetic responses to elevated pCO 2 is uncertain and variable (Koch et al. 2013).Mercado & Gordillo (2011) re ported that changes in CO 2 concentration by natural processes or climate change could have a limited impact on primary production in a variety of aquatic ecosystems due to the effective acclimation processes (Beardall & Raven 2004).
In the present study, a reduction in photosynthetic capacity under the short-term incubation at elevated pCO 2 was observed for C. tamariscifolia, while P. pavonica increased its net photosynthesis.These observed changes in net photosynthesis were not followed by changes in ETR max , nor in other chl a fluorescence parameters.In fact, an increase in F v /F m was detected under high pCO 2 in C. tamariscifolia.By comparing the results of several studies, one can say that the effects of OA on photosynthesis of macroalgae varies according to species and functional traits, as well as incubation characteristics (time exposure, light quality and quantity, type and size of incubator, etc.) (Martin & Gattuso 2009, Semesi et al. 2009, Gao & Zheng 2010, Sinutok et al. 2011, Zou et al. 2011).Important physiological aspects to be considered are the presence of an HCO 3 − transport system, the type of carbon-concentrating mechanisms (CCM) (Mercado et al. 1998, Wu et al. 2008) and the possible inhibitory effect of CO 2 on respiration, among others.CCMs involve the enzyme external carbonic anhydrase (exCA).A previous study showed that exCA was not present in P. pavonica, but it was detected in C. tamariscifolia, indicating a low affinity of this latter species for inorganic carbon compared to other macroalgae (Mercado et al. 1998).An exCA in P. pavonica could be present, but Mercado et al. (1998) did not detect it as consequence of the limitation in the methodology used to detect this enzyme.For another Padina species, P. sanctae-crucis, Enríquez & Rodríguez-Román (2006) suggested the presence of an efficient CCM, probably related to HCO 3 − uptake.A CCM requires an energetic investment for ex pression and operation (Raven et al. 2012).A down-regulation of the CCM activity in response to en riched inorganic carbon has been proposed; therefore, the alga has more energy to invest in other processes such as growth (Giordano et al. 2005).In spite of these mechanisms, we did not find any positive effect of elevated pCO 2 on photosynthesis in the noncalcifying macroalgae C. tamariscifolia.
In this study, E. elongata showed the lowest pigment concentration among the 3 species (especially fucoxanthin and violaxanthin contents which were half those in the brown algae), although antheraxanthin was maintained in the same order of magnitude.To our knowledge, no short-term study (hours) on the effect of high CO 2 on pigment content exists.In this study, no significant effect of pCO 2 treatments in pigment content was found in the non-calcifying C. tamariscifolia or in the calcifying P. pavonica.Nevertheless, P. pavonica showed a tendency to have higher pigment contents under high CO 2 , especially fucoxanthin, which is an accessory photosynthetic pigment with strong antioxidant properties (Mori et al. 2004, Fung et al. 2013)  In response to intense solar radiation, algae have evolved certain photoprotective mechanisms by accumulating a series of photoprotective compounds, such as carotenoids, phenolic compounds (brown algae) and MAAs (red algae).Regarding phenolic compounds, a higher content was observed in C. tamariscifolia than in P. pavonica.It is possible that the acidification of the medium drove the phenolic losses in C. tamariscifolia and P. pavonica by photo degradation and release, as has been observed in marine angiosperms along a natural gradient of CO 2 (Arnold et al. 2012).A decrease of phenolic compounds and the antioxidant activity during a submarine volcanic eruption, which produced a decrease in seawater pH level, was also observed (Betancor et al. 2014).The higher concentration of zeaxanthin observed in E. elongata under high CO 2 treatment could indicate a higher photoprotection potential under a future scenario of high pCO 2 .Hence, a photo protective role could be argued for zeaxanthin, since it has been described as a zeaxanthin-dependent amplification of NPQ exclusively found in thylakoids containing zeaxanthin (Goss et al. 2006).To our knowledge, this is the first time that the effect of different pCO 2 on MAAs has been studied.Al though no effect on total MAA content was found in E. elongata, a tendency to accumulate relatively more palythine under high CO 2 was ob served.It is known that palythine possesses a higher antioxidant capacity than the other MAAs (De la Coba et al. 2009).Thus, even though this species may show a loss of carbonate skeleton under OA, it could still maintain high photoprotective and antioxidant capacities.
The ecophysiological responses of macroalgae to high CO 2 concentrations are variable and complex, and for calcareous macroalgae, the responses may be even more complex due to the calcification process.Some evidences show that certain calcifying Phaeophyceae could be amongst the ecological winners under OA scenarios (Kübler et al. 1999, Porzio et al. 2011, Raven 2011, Johnson et al. 2012), and that the function and structure of future ecosystems could be drifting towards these species.However, the ecological impacts on these particular species and further consequences for the surrounding macroalgal community are unknown and difficult to predict.

CONCLUSIONS
In this study, we present a novel experimental design to incubate macroalgae in situ at different pCO 2 .Our design tested non-calcifying (Cystoseira tamariscifolia) versus calcifying (Padina pavo nica and Ellisolandia elongata) marine macroalgae in a short-term incubation at 2 different pCO 2 : air (ambient CO 2 ) and the p CO 2 predicted by the end of the 21st century (700 µatm, high CO 2 ).Slight differences were detected in the 3 studied species after the shortterm incubations.Although one would expect an increase in net photosynthesis under enriched CO 2 as the algae would have more substrate for Rubisco, slight differences were found between CO 2 levels within in vivo chlorophyll fluorescence parameters and a positive effect of increasing pCO 2 on net photosynthesis was only observed in P. pavonica.On the other hand, although the activation of the NPQ dissipation mechanism occurred under high CO 2 in the brown algae P. pavonica (indicating a high protection mechanism), this capacity was counteracted by a loss of phenols.In E. elongata, the higher zeaxanthin and palythine contents under high CO 2 could indicate a higher photoprotection capacity.This is one of the few studies in which the effect of CO 2 on macroalgal photoprotective compounds has been evaluated (but see Betancor et al. 2014).
Despite the increasing number of studies on the effect of changes in pCO 2 on macroalgae, the ecophysiological responses of these species to a future scenario of OA are still unknown.Together with the other systems proposed for different habitats and macrophyte species (Campbell & Fourqurean 2011, 2013, Kline et al. 2012, Arnold et al. 2012), the in situ incubation system proposed here may contribute to this knowledge and has the advantage of being simple, reproducible and cheap.Nevertheless, for the near future, parameters such as different pCO 2 and gas flux rates, carbonate system parameters, chamber volume and algal density or biomass (among others) should be exhaustively controlled.The development of long-term experiments including day−night cycles and enabling acclimation responses should be studied for a better understanding of how macroalgae will respond to a future OA scenario.tribution to the GAP 9 workshop 'Influence of the pulsedsupply of nitrogen on primary productivity in phytoplankton and marine macrophytes: an experimental ap proach'.The authors especially thank the Office of the Cabo de Gata-Níjar Natural Park (Consejería de Medio Ambiente y Ordenación del Territorio, Junta de Andalucía) for their help and for giving us the opportunity to work in the Natural Park.We thank the reviewers for their helpful and con structive comments, which have significantly improved the manuscript.

Fig. 4 .
Fig. 4. Phenolic compound tissue concentration in Cystoseira tamariscifolia and Padina pavonica after 5.5 h in situ incubation in high CO 2 and ambient CO 2 .Data are expressed as mean + SE (n = 12).Lowercase letters denote significant differences

Table 2 .
ANOVAs of the effect of CO 2 treatments on alkalinity, temperature, and pH for Cystoseira tamariscifolia, Padina pavonica and Ellisolandia elongata.Significant results (ANOVA, p < 0.05) indicated in bold

Table 4 .
Maximal quantum yield (F v /F m ), maximal electron transport rate (ETR max ), photosynthetic efficiency (α ETR ), saturation irradiance for ETR (Ek ETR ), maximal non-photochemical quenching (NPQ max ), the slope of the NPQ versus irradiance (α NPQ ) and the saturation irradiance for NPQ (Ek NPQ ) of Cystoseira tamariscifolia, Padina pavonica and Ellisolandia elongata after 5.5 h in situ incubation in high CO 2 and ambient CO 2 conditions.Values are mean ± SE (n = 12).Lowercase letters denote significant differences (ANOVA, p < 0.05) environmental conditions and deal with the variation of the natural populations(Wernberg et al. 2012).In this study, macroalgae were incubated in situ at 2 different pCO 2 levels, while other para meters remain relatively unchanged.The experiments were performed in a pristine marine environment, by evaluating the short-term responses of Cystoseira tama riscifolia, Padina pavonica and Ellisolandia elongata, which are dominant macroalgae on the Mediterranean shores.

Table 5 .
. Although the pH values were significantly lowered (by 0.3 units) in the cylinders treated with high CO 2 Pigment concentration of Cystoseira tamariscifolia, Padina pavoniva and Ellisolandia elongata after 5.5 h in situ incubation in high CO 2 and ambient CO 2 conditions.Values are mean ± SE (n = 12).Chl a, chl c, phycoerythrin and phycocyanin are in mg g −1 DW.The other pigments are expressed as µg g −1 DW.Lowercase letters denote significant differences (ANOVA, p < 0.05) and its presence strongly enhanced cell viability against H 2 O 2 induced oxi dative damage(Heo et al. 2008).Additionally, both brown algae species showed increases in NPQ para meters (NPQ max and α NPQ , for C. tamariscifolia and P. pavonica, respectively) and decreased Ek ETR under high CO 2 .The lower Ek ETR indicates a higher rate of reactive oxygen species (ROS) production at incubation irradiances (solar radiation)(Lesser 2006).It is pos sible that the short-term exposure to high CO 2 activated their heat dissipation mechanisms; hence, these species became protected against the in creased photo oxidative damage (Demmig-Adams & Adams 2006).