Inorganic Carbon Acquisition and Photosynthetic Metabolism in Marine Photoautotrophs: A Summary

1. Introduction

Although oceans have supported oxygenic photosynthesis of cyanobacteria for close to 3.5 billion years, and about half of earth’s primary (photosynthetic) production today occurs in our oceans [1], the study of the mechanisms involved in marine photosynthesis lags far behind that of terrestrial plants, which colonized land ‘only’ some 400 million years ago. This is despite the fact that the discovery and elucidation of the (almost) universal pathway of CO₂ fixation and reduction, the Calvin-Benson-Bassham cycle (or the photosynthetic carbon reduction cycle, PCRC), was based on work using an, albeit freshwater, unicellular alga.

Two distinct differences between photosynthesis in terrestrial and aquatic environments are that 1) the diffusivity of solutes, including CO₂, is orders of magnitude lower in water than in air, and 2) irradiance is exponentially attenuated with depth such that the euphotic zone often extends only a few tens of metres down along a depth gradient [2]. Regarding photosynthetically active radiation (PAR), the dogma has been that net photosynthesis (photosynthetic activity corrected for mitochondrial respiration and other, minor, carbon losses) on a diel cycle is positive (and thus supports growth) down to a depth where 1% of surface irradiance remains. However, it is now apparent that there are many exceptions to this, where net photoautotrophic growth occurs at much lower light levels [3,4]. Marine angiosperms (seagrasses, which colonised the seas secondarily only less than 100 million years ago) must support an often-extensive rhizome and root biomass and, thus, need up to 10% of surface light in order to grow (Chapter 9 in [5]). While adaptations to various environments, including the various and varying irradiances, are a must in both terrestrial and marine photoautotrophs (see [1,2,5]), we will summarise here our current understanding of what we see as the main difference between the two groups, i.e. how the marine species in general, as well as using specific examples, cope with the scarce supply of CO₂ in seawater while also featuring a generally low CO₂-affinity of their ultimate carboxylating enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco).

2. Marine Inorganic Carbon Sources for Carbon Assimilation

When in equilibrium with today’s atmospheric CO₂ concentration of ~410 ppm (or 18 μM), and rising, the marine dissolved CO₂ concentration at 20^oC and a salinity of 35 is some 25% lower (a value which depends on temperature and salinity). In addition, the diffusivity of solutes (including CO₂) in water is orders of magnitude slower than in air. Therefore, and given the generally low CO₂-affinity of Rubisco in marine photoautotrophs (see below), photosynthetic rates of the latter would be severely limited were they to depend on dissolved CO₂ only. However, atmospheric CO₂ (CO_2air) reacts with water molecules to form carbonic acid (H₂CO₃), which dissociates to form bicarbonate (HCO₃^-) and carbonate (CO₃^2-) ions according to Equation 1.

CO_2air ↔ CO₂ + H₂O ↔ H₂CO₃ ↔ HCO₃^- + H⁺ ↔ CO₃^2- + H⁺ (Eqn. 1)

18 μM 14 μM CA <1 μM ~1,700 μM ~300 μM

Air │ Seawater (at pH 8.1, 20⁰C and a salinity of 35)

As can be seen, the ionic concentrations in equilibrium with dissolved CO₂ are strongly dependent on pH (H⁺), and for most submerged marine photoautotrophs the most available Ci form (in terms of concentration) is HCO₃^-, which e.g. in today’s air-equilibrated seawater pH of 8.1 is 120 times higher than that of CO₂. (CO₃^2- is not used in marine photosynthesis.) Secondly, the hydration of CO₂ to form H₂CO₃ and vice versa is a slow process with half-times approaching a minute. However, it becomes virtually instantaneous when catalysed by the enzyme carbonic anhydrase (CA), and the importance of this enzyme in marine photosynthetic Ci acquisition will become apparent below.

3. Assimilation of Inorganic Carbon via Rubisco

The central enzyme of carbon assimilation in cyanobacteria, eukaryotic algae, marine (seagrasses) and terrestrial vascular plants, is Rubisco, which is the first step in the PCRC for net assimilation of inorganic carbon to organic matter. Although other pathways for the conversion of CO₂ to organic matter exist, those are found in non-oxygenic autotrophic bacteria [6]. The reactions of the PCRC and differences in its regulation among cyanobacteria and eukaryotic algae have been recently discussed in [7].

Rubisco will catalyse both the carboxylation of ribulose bisphosphate, to yield 2 molecules of phosphoglycerate), and its oxygenation to yield one phosphoglycolate plus one phosphoglycerate according to Equations 2 and 3 below:

CO₂ + H₂O + D-ribulose-1,5-bisphosphate → 2 X 3-phosphoglycerate (Eqn. 2)

O₂ + D-ribulose-1,5-bisphosphate → 3-phosphoglycerate + 2-phosphoglycolate (Eqn. 3)

The phosphoglycolate compound) produced by the reaction shown in Equation 3 can be processed by the reactions of photorespiration, which sees the loss of one of the carbons as CO₂, with the other carbon being recouped as glycerate. It has been shown [8] that, at least in the cyanobacterium Synechococcus, this is achieved through three pathways, i.e. 1) a ‘conventional’ PCOC involving glycolate dehydrogenase, and regeneration of glycerate via glyoxylate, glycine, serine and hydroxypyruvate, 2) the glycerate via tartronic semialdehyde, or 3) conversion of glyoxylate via oxalate and formate, to CO₂. Further details of photorespiration in aquatic phototrophs can be found in [9].

Since the two reactions catalysed by Rubisco are competitive, using the same active site on the enzyme, the relative rate of the two reactions in a photosynthesising cell can be given by the specificity factor S_rel according to Equation 4,

S_rel=[K_0.5(O₂)•k_cat(CO₂)]/[K_0.5(CO₂)•K_cat(O₂)] (Eqn. 4)

where k_cat(CO₂) = CO₂ saturated specific rate of carboxylase activity of RUBISCO (mol CO₂ mol^-1 active sites s^-1), K_0.5(CO₂) = concentration of CO₂ at which the CO₂ fixation rate is half of k_cat(CO₂), k_cat(O₂) = O₂ saturated specific rate of oxygenase activity of RUBISCO (mol O₂ mol^-1 active site s^-1) and K_0.5(O₂) = concentration of O₂ at which the O₂ fixation rate is half of k_cat(O₂). The lower the selectivity factor, the higher will be the impact of oxygenation on net photosynthetic CO₂ fixation and photorespiration. For organisms with the lowest S_rel values, the kinetics of Rubisco do not allow for net fixation of CO₂ at air-equilibrate seawater (or freshwater), irrespective of the pathway of glycolate metabolism used in photorespiration [9].

4. Kinetics of Rubisco vs. Whole Cell Photosynthesis

4.1. Cyanobacteria and Microalgae

The forms of Rubisco vary considerably across the cyanobacterial and microalgal groups and show different kinetic properties and selectivity towards CO₂ and O₂. The general trend across all photoautotrophs, though, is that a low K_0.5(CO₂) and a high S_rel are correlated with a low k_cat(CO₂) and vice versa. Cyanobacteria tend to possess Rubiscos with high K_0.5(CO₂) and somewhat lower selectivity factors than the eukaryotic algae, which may reflect the cyanobacterial evolutionary origins when CO₂ levels were higher and O₂ concentrations lower than the present-day values [10,11]. Affinities of Rubiscos from eukaryotic algae tend to be higher (lower K_0.5(CO₂)) and S_rel values higher. Values of kinetic properties of Rubiscos from a range of cyanobacteria, and microalgae are given in Table 1. This table includes values from both marine and freshwater species, but differences in values reflect more the taxonomic position of the species than their salinity environments.

We note that the values of K_0.5(CO₂) for Rubisco are in most cases higher than the air-equilibrium CO₂ concentrations generally found in aquatic environments (10-25 μM, depending on temperature, salinity, and the presence and density of various respiring and photosynthesising organisms, see Section 7). Consequently, air-equilibrium levels of CO₂ would, all other things being equal, be limiting to carbon assimilation by cyanobacteria and microalgae. Furthermore, most cyanobacteria and microalgae show much higher affinity for Ci in whole cell photosynthesis than is exhibited by the isolated Rubisco. To cite a few examples, values for K_0.5(CO₂) in Ci-dependent photosynthesis of cells of Phaeodactylum tricornutum have been reported as 0.53-4.5 μM [12,13] whereas that for Rubisco is 28-41 μM (see [14] and references therein). Similarly, while [15] report that Chaetoceros muelleri cells has K_0.5(CO₂) values of ~0.26 μM in photosynthesis, [16] quote an equivalent value for Rubisco from this species of 23 μM. The marine cyanobacterium Synechococcus PCC7002 with Form 1Bc Rubisco has a K_0.5(CO₂) for Rubisco of 185 μM [14] and the equivalent value for photosynthesis of ~0.7 μM [17]. There are only a few exceptions to this trend, with the most rigorous examples coming from freshwater Synurophyceae, which have evolved Rubiscos with a high affinity for CO₂ ranging from 18-42 μM [18] compared to similar values of whole cell photosynthesis of 22-45 μM.

It is clear that cyanobacteria and most microalgae have evolved mechanisms to enhance CO₂ concentrations at the active site of Rubisco such that carbon assimilation is at, or close to, CO₂-saturation at air-equilibrium levels in the surrounding medium. These CO₂-concentrating mechanisms (CCM) are discussed below in Section 5.

Table 1. Kinetic properties of Rubiscos from different cyanobacterial and microalgal groups. Note these data include values from freshwater as well as marine species. After [9,14] and references therein.

Organisms	Rubisco	K0.5 CO2 (mM)	Srel (mol mol−1)	Kcat (mol CO2 mol−1 active sites s−1)
b-cyanobacteria	Form IBc	200-260 *	35-56	2.6 - 11.4
a-cyanobacterium Prochlorococcus	Form 1Ac	750		4.7
Green algae	Form 1B	29-38	61-83
Diatoms	Form 1D	23-68	57-116	2.1-3.7
Synurophyceae	Form 1D	18.2-41.8
Dinoflagellates#	Form II		~37
Rhodophyta Porphyridium purpureum	Form 1D	22	144	2.6
Rhodophyta Cyanidium	Form 1D	6.6-6.7	224-238	1.3-1.6

* A minority of reports have lower values down to 105 mM. ^#Kinetic data on dinoflagellate Rubiscos is limited due to the instability of their Form II Rubiscos.

Table 1. Kinetic properties of Rubiscos from different cyanobacterial and microalgal groups. Note these data include values from freshwater as well as marine species. After [9,14] and references therein.

Organisms	Rubisco	K0.5 CO2 (mM)	Srel (mol mol−1)	Kcat (mol CO2 mol−1 active sites s−1)
b-cyanobacteria	Form IBc	200-260 *	35-56	2.6 - 11.4
a-cyanobacterium Prochlorococcus	Form 1Ac	750		4.7
Green algae	Form 1B	29-38	61-83
Diatoms	Form 1D	23-68	57-116	2.1-3.7
Synurophyceae	Form 1D	18.2-41.8
Dinoflagellates#	Form II		~37
Rhodophyta Porphyridium purpureum	Form 1D	22	144	2.6
Rhodophyta Cyanidium	Form 1D	6.6-6.7	224-238	1.3-1.6

* A minority of reports have lower values down to 105 mM. ^#Kinetic data on dinoflagellate Rubiscos is limited due to the instability of their Form II Rubiscos.

4.2. Macroalgae and Seagrasses

The K_0.5(CO₂) of Rubiscos in macroalgae varies between 30 and 85 μM. This was established for several common macroalgae from all three phyla such as the green algae Ulva expansa, Ulva fasciata, Enteromorpha (now Ulva) linza, the brown Fucus gardneii, Padina pavonia, Dictyota fasciola, and the red algae Polyneura latissima, lridaea cordata, and Gigartina latissimi [19,20,21,22,23,24]. In all cases, we conclude that the CO₂ concentration giving half-maximal rates of Rubisco carboxylation in marine macroalgae are higher than the marine concentration of CO₂ and, so, they require a CCM in order to perform optimal rates of photosynthesis. These K_0.5(CO₂) values when measured under low O₂ levels were not significantly higher than when measured in air [23], indicating a more efficient carboxylation via Rubisco than in seagrasses (see below) as well as a virtual lack of photorespiration [21].

While the different K_0.5(CO₂) values can partly be due to the red and brown macroalgae possessing the Form 1D while the green algae and seagrasses contain Form 1B Rubiscos [25], the different values for species which have the same form is less understood. However, the link between the evolutionary occurrence of various Rubiscos and the evolution of marine photoautotrophs has been reviewed by [26] and has gained much recent interest [25].

Seagrasses are marine angiosperms that colonised shallow softbottom areas some 100 million years ago. Like their terrestrial counterparts from which they evolved, they possess roots and rhizomes as well as flowers and pollination (although meadows mostly grow clonally by rhizome extensions). The photosynthetic traits of these plants have been reviewed in [27,28,29] and, more recently, in [25]. K_0.5(CO₂) values for Rubiscos in seagrasses have recently been measured to be around 40 μM with values being significantly lower in the absence of O₂ [25], indicating the presence of photorespiration. These include the Mediterranean species Posidonia oceanica (K_m(CO₂) ~45 μM), Cymodocea nodosa (37 μM) and the temperate Zostera marina (32 μM) under air-equilibrium O₂ concentrations. As is the case for macroalgae, these plants must thus also possess CCMs.

We conclude in this section that the low availability of CO₂ in seawater and the low affinity of Rubiscos for CO₂ necessitates the presence of a CCM in almost all marine photoautotrophs. As follows, we will see that such CCMs are most often based on the use of HCO₃^- which, again, is present at a much higher concentration in seawater than CO₂.

5. Inorganic Carbon Acquisition and ‘Biophysical’ CCMs

Measurements of the kinetics of whole-cell photosynthesis vs. CO₂ in most marine phototrophs show K_0.5(CO₂) values that are much lower than for the isolated Rubisco. This higher affinity of intact cells/tissues implies that the CO₂ concentration at the active site of Rubisco is higher than in the bulk medium, i.e. a CCM is present. This is supported by direct measurements of internal CO₂ pools with CO_2in vs. CO_2out ratios ranging from ~2 fold in symbiotic dinoflagellates, 20-fold in the macroalga Ulva fasciata, to 800-900 in cyanobacteria, though there is considerable variation depending on species and environmental conditions (see Table 1 in [9] and references therein). There is also other, indirect, evidence for CCM activity including carbon isotope discrimination values against ¹³C being much less negative than for isolated Rubisco, low or absent inhibition of photosynthesis by O₂ and low CO₂ compensation points in photosynthesis vs. Ci experiments.

Inorganic carbon assimilation by marine phototrophs is, in most cases, driven by the use of HCO₃^-, the main Ci source in seawater. Bicarbonate acquisition can be demonstrated in several ways. One approach, which may also be the easiest one, is to measure final pH values in pH-drift experiments [30]. In short, cell suspensions, macroalgae or seagrasses (or parts thereof) are inserted into a closed illuminated chamber in which the rise in pH is recorded as they photosynthesise. If the final pH is around 9 or higher, then this is taken as evidence for HCO₃^- use since the CO₂ concentration is extremely low at the high pH values generated in closed systems. If, on the other hand, the final pH is 8.5 or less, then HCO₃^- is not used since this ion is present but obviously does not contribute to raise the pH by photosynthesis. Thus, those (albeit few) algae and plants would be CO₂-only users. Burns and Beardall [12] used the pH dependence of K_0.5CO₂ and independence of K_0.5(HCO3^-) to infer HCO₃^- use in a number of marine microalgae. Another way to demonstrate HCO₃^- use is to determine the ¹³C vs. ¹²C contents in algal and plant material vs. those of the Ci sources in the seawater. These values are expressed as Δ¹³C (δ¹³C if the source ¹³C:¹²C ratio is not determined). Isolated Rubisco shows values close to -27‰ to-30‰ but HCO₃^- use and activity of a CCM is reflected in less negative values. A summary of the use of δ¹³C in investigations of marine photoautotrophs may be found in [31]. Isotope disequilibrium techniques as described by [32,33] can also be used to indicate HCO₃^- vs. CO₂ use. Alternatively, membrane inlet mass spectroscopy as described for marine Synechococcus [17] can be used to directly determine rates of uptake of CO₂ and/or HCO₃^-.

5.1. Cyanobacteria and Microalgae

Use of HCO₃^- and its role in driving CCM activity has been demonstrated in a range of cyanobacteria and microalgae, and CO₂ can also cross the plasma membrane by diffusion, in some cases aided by aquaporins [7,9,34].

In cyanobacteria there are multiple systems for active Ci entry into cells and accumulation of CO₂ within the carboxysome, with some differences in the transporters between marine and freshwater species [35]. Most of these transport systems involve direct active transport of HCO₃^- or an energised conversion of CO₂ to HCO₃⁻ via a NAD(P)H dehydrogenase, which effectively acts like a Ci-pump, even though direct active transport of CO₂ has not occurred ([34] and references therein, [35]). Enhanced HCO₃^- supply to the carboxysome (the site of Rubisco activity in cyanobacteria) leads to increased CO₂ levels, involving the activity of a carboxysomal CA (see [34] and references therein).

The CCMs in eukaryotic microalgae are less well elucidated than those of cyanobacteria. Nonetheless, HCO₃^- transporters (SCL4-type transporters) associated with the plasma membrane have been identified in the marine diatoms Phaeodactylum tricornutum and Thalassiosira pseudonana [36]. Energy for this process appears to be associated with linear electron flow involving both photosystems II and I [37]; in eustigmatophyte algae respiratory derived ATP is used [38]. There is also evidence, based on observations for some species, of active transport of CO₂ and HCO₃⁻ and CO₂ accumulation by isolated chloroplasts [39] and there is evidence for SCL4-type transporters at the chloroplast envelope of diatoms [36,40] and a range of transporters of inorganic carbon have been postulated for all four of the plastid envelope membranes [41].

In eukaryotic microalgae (but not cyanobacteria), Ci acquisition can be modulated by an extracellular CA associated with the cell wall. This converts HCO₃^- to CO₂ in the periplasmic space, thereby increasing CO₂ concentrations locally and assisting its diffusion across the cell wall and plasma membrane. A range of other CAs are present in various compartments within the eukaryotic algal cell and are involved in maintaining the equilibrium between CO₂ and HCO₃^- [34,42].

5.2. Macroalgae and Seagrasses

The macroalgal photosynthetic traits studied the most are for various species of Ulva (reviewed in [43]), and this genus will therefore be used here as an example applicable for many others that have been studied, as well as what will most probably be found in future studies. So, among other genera of red, brown and green algae, Ulva spp. (including what was once named Enteromorpha) stand out as being extreme HCO₃^- users, i.e. a macroalga in which additional CO₂ or Ci beyond the normal seawater Ci composition will not significantly increase photosynthetic rates ([44,45] and Figure 7.5 in [5]). Species of this genus can use HCO₃^- in two different ways (see below) and the Rubisco K_0.5(CO₂) for subtropical Ulva fasciata was found to be 70 μM [24]. In comparison with today’s open-water CO₂ concentration of 14 μM, at the salinity and temperature used, it can be ascertained that a diffusive CO₂-equilibrium could yield only ~5% of this alga’s photosynthetic potential as based on its Rubisco – if assuming that the CO₂ concentration inside the photosynthesising cells would be in equilibrium with that of the seawater (which is an overestimation). Thus, like most macroalgae, Ulva must possess a CCM.

A spin-off from pH-drift experiments is to actively vary the pH with HCl or NaOH within an illuminated closed system containing algae or seagrasses (or, usually, parts thereof) while measuring both pH and photosynthetic rates, usually by O₂ evolution. Given the total Ci in natural seawater (~2.2 mM) and the pH, the Ci composition can then be calculated. The results of experiments using such systems showed already early on that the common macroalgal genus Ulva used HCO₃^- in such an efficient way that this Ci form saturated photosynthesis without any need for supplemental Ci ([45,46] and Figure 7.5 in [5]). This was not the case for seagrasses where both subtropical [47] and, later, temperate and Mediterranean [48,49,50] species could use HCO₃^- under normal seawater Ci conditions, but increased their photosynthetic rates, often more than doubling them, as CO₂ was released by lowering the pH to <8 (see Figure 7.12 in [5]). Our conclusion of the above-mentioned approaches is that marine macrophytes in general use HCO₃^- as their external Ci source for photosynthesis, but that seagrasses do so less efficiently and, thus, are Ci-limited in today’s air-equilibrated seawater and will benefit from future increases in atmospheric and, consequently, dissolved CO₂ levels. This general conclusion was somewhat challenged recently [11] as being an underestimation of the capacity of seagrasses to photosynthesise at close to the HCO₃^- concentration of today’s oceans (see also Section 7 below).

Two cautionary notes before we continue: Most experiments on photosynthetic properties of marine photoautotrophs are done in laboratories. For the macrophytes, this usually means that parts of thalli or seagrass leaves are cut into sections so as to fit into small O₂-electrode vials. Such (mis)handlings may show results that are different from those measured in situ. This was shown for the shallow-growing tropical seagrasses Halophila ovalis and Cymodocea serrulata, in which results of previous laboratory experiments showed that they were Ci-limited in the ~2.2 mM-Ci normal seawater [51,52]. However, when the measurements were done in situ on intact plants, both species were Ci-saturated [47]. Secondly, buffers have often been used in order to keep a certain pH while measuring O₂ evolution. However, it was shown that TRIS buffer itself lowered photosynthetic rates even at normal seawater pH and Ci conditions [53]. This led to the conclusion that e.g. the common temperate seagrass Zostera marina was highly dependent on active trans-membrane pumps producing protons (H⁺) toward the DBL which, when using buffers, were neutralised.

As mentioned before, the most common way for marine photoautotrophs to utilise HCO₃^- is to convert it to CO₂ within their diffusion boundary layers (DBL, including the cell walls) as catalysed by membrane-bound, extracellularly acting, CA (Figure 2 in [43]). If the pH within the DBL where CA activity is present is lower than that of the surrounding seawater (e.g. in acid zones [29], then this will account for an efficient conversion of HCO₃^- to high concentrations of CO₂ that can diffuse through the plasma membrane into the photosynthesising cells. This way of Ci utilisation in macroalgae and, especially, in most seagrasses, is today common knowledge that has been detailed in e.g. Chapter 7 in textbook [5] and reviews [11,27,28,29]. The easiest way to detect this form of HCO₃^- use is to add the membrane-impermeable CA inhibitor acetazolamide (AZ) while measuring photosynthesis; if rates cease, then this is a strong indication of this mechanism being used. If not, then either another way to use HCO₃^- (see the following) is in effect, or CO₂ only (Section 6) is used.

The most efficient “other way” to use HCO₃^- from seawater is by its direct uptake (Figure 3 in [43]). Being an ion, HCO₃^- cannot easily diffuse through plasma membranes and, so, needs to be transported. For subtropical Ulva sp. it was found that photosynthesis was fully inhibited by 4,4’-diisothiocyanostilbene-2,2’disulphonate (DIDS) [54] which, till then, had been used as a classical inhibitor of HCO₃^- transport in red blood cells (RBC) via a membrane-bound anion exchange (AE) protein. It was further found that the AE from Ulva and from RBCs shared very similar properties [54,55] with the main difference that RBCs exchanged HCO₃^- for chloride (Cl^-) while Ulva exchanged it for hydroxyl ions (OH^-) [56]. Interestingly, when AZ-sensitive temperate Ulva lactuca was exposed to pH 9.8 for some 10 h, it converted to a direct, DIDS-sensitive, HCO₃^--uptake alga similar to those Ulva spp. growing in subtropical waters of the Mediterranean [57]. When in this state, Ulva had a 10 times higher affinity for HCO₃^- than the one growing in temperate waters. It was concluded that, ultimately, the pH near the plasma membrane was the trigger for the alternative ways in which subtropical Ulva fasciata and temperate Ulva lactuca used HCO₃^-. The addition of DIDS was subsequently used by several researchers as a way to indicate direct uptake of HCO₃^- so, besides Ulva, Chaetomorpha melagonium [58], Macrocystis pyrifera [59] and Posidonia oceanica [60] could also transport HCO₃^- into their photosynthesising cells.

There is, of course, an energy cost for CCMs [61] which obviously is met by the irradiances where CCM-requiring algae and seagrasses grow. At lower light (at depts) and in colder waters, some macroalgae may not need CCMs and can, like terrestrial C₃ plants, use the, albeit slow, diffusional supply of CO₂ for their lower photosynthetic rates. These algae include several rhodophytes [62], but they still represent an exception among macroalgae which, again, use HCO₃^-. The same goes for algae growing in subarctic or Arctic areas [63] where photosynthetic rates are restricted by low temperatures (see however a report on two polar red algae being HCO₃^- users and possessing CCMs [64].

Several parameters found in macroalgae form strong evidence that they a) use HCO₃^- as their external Ci source under today’s marine Ci composition [30], and b) that they possess CCMs. These include low CO₂-compensation points for many macroalgae [30,65] including the common species Enteromorpha (now Ulva) intestinalis, Ulva lactuca, Porphyra umbicalis, Palmaria (then Rhodymenia) palmata), Fucus serratus, and Pelvetia canaliculata; in all cases were the CO₂ compensation concentrations well below ~10 μM, showing that virtually no CO₂ is leaked outward even at low CO₂ concentrations and, thus, indicating an efficient Ci utilisation system (a CCM) that keeps all CO₂ within the photosynthesising cells to be used by photosynthesis. There is, however, a lack of direct evidence that intracellular CO₂ concentrations surrounding Rubisco are higher than that of the surrounding seawater. As far as we know, such estimates are limited to, again, Ulva fasciata where intracellular Ci concentrations within the photosynthesising cells were found to be 220 μM [66]. With a Rubisco K_m(CO₂) of 70 μM, and assuming an intracellular pH of 7.2 [67] and that ~200 μM CO₂ was present also within the chloroplasts, this meant that this alga could concentrate CO₂ so as to fully saturate its Rubisco.

A positive correlation between deep- and shallower-growing (<10 m) macrophytes (macroalgae and seagrasses), as based on δ¹³C values and end-pH values of pH-drift experiments indicates that the deep-growing ones are less in need for CCMs (see Section 6 below).

Seagrasses in general increase their rates of photosynthesis when CO₂ is allowed to increase above air-equilibrated seawater concentrations (~14 μM today). This can be done by lowering the pH or adding Ci in a closed system [52,54], or by adding Ci above the natural seawater concentration ([11,48,49,68], see Figure 7.12 in [5] for an early example). There is a whole range of affinities to Ci in various species; some seem to be severely limited in their HCO₃^- utilisation capacity (e.g. Thallasia testudinum) while others are more efficient (e.g. Halodule wrightii, Syringodium filiforme [48], Posidonia oceanica and are less affected by additional Ci or CO₂ [49,50].

Rubisco originally evolved when CO₂ levels were very much higher, and O₂ levels lower, than in the present day; most probably during periods when CO₂ fell to 2–16 times the present atmospheric level, depending on Rubisco kinetics [10,26]. Adaptations of Rubisco’s capacities, and how it co-evolved with photoautotrophs, was recently thoroughly reviewed [11] and will not be treated further here.

There have been a few additional reviews on Ci utilisation and CCMs specifically in seagrasses [28,29,69], and, lately, [11], all of which generally agree on HCO₃^- being the main external Ci form used.

Some variations on the theme of external-CA mediated HCO₃^- use in some macroalgae and seagrasses have also been reported. One originated in 2001 from experiments in which buffers (usually TRIS) were (mis)used for keeping certain pH values within O₂ electrode systems. It was then found that the TRIS buffer itself at the natural seawater pH had a negative effect on photosynthetic rates of the seagrass Zostera marina [53]. This led to the assumption that H⁺ release was an alternative way for this seagrass to utilise HCO₃^- by neutralising H⁺ efflux, which otherwise would co-transport HCO₃^- into the photosynthesising cells [69] (see (b) in Figure 7b in [5]. Another way of using HCO₃^- was also suggested for the Indo-Pacific seagrass Halophila stipulacea in which both TRIS and AZ together could impede CO₂ formation and fluxes of CO₂ into the cells ((c) in Figure 7.13 in [5]). This was also found for several warmer-water species such as Halodule wrightii, Halophila ovalis, and Cymodocea rotundata [70]. However, most works points to the external CA-mediated conversion of HCO₃^- to CO₂ as the commonest way of Ci acquisition in seagrasses (Figure 7.7 and (a) in Figure 7.13 in [5]).

Photorespiration (the beginning stage of which is given in Equation 3) can be present in plants lacking efficient CCMs. This is true for e.g. the seagrass Zostera marina and, often seen as a wasteful process, can lower photosynthetic rates significantly, especially under conditions of low Ci and high O₂ levels such as generated when growing together with the highly efficient photosynthesiser Ulva ([71] and see Section 7). On the other hand, photorespiration can also protect the seagrass from photodamage by dissipating solar energy at high irradiances. The presence of photorespiration was indicated by a lower gross O₂ evolution rate under natural O₂ conditions than when O₂ was reduced.

6. Alternative ‘Biochemical’ Modes of Inorganic Carbon Utilisation in Some Marine Photoautotrophs

The CCMs described above are sometimes referred to as ‘biophysical’ CCMs because they involve active or facilitated transport of Ci into the photosynthesising cells. Although uncommon, there have been reports of phosphoenolpyruvate carboxylase (PEPC) and phosphoenolpyruvate carboxykinase (PEPCK) being primary carboxylases present in micro- and macro-algae and leading to ‘biochemical’ CCMs similar to those found in terrestrial C₄ and CAM plants.

6.1. Cyanobacteria and Microalgae

Despite early reports of C₄-like photosynthesis in a marine diatom [72], later work [73] ascribed the observed labelling patterns and enzyme activities to high rates of anaplerotic β-carboxylation reactions, necessary to ‘top up’ intermediates of the TCA cycle. While single-cell C₄ photosynthesis has been more recently reported by [74,75], it is now more generally accepted that among the microalgae this is only present in the diatom Thalassiosira weissflogii, which shows C₃–C₄ intermediate Ci assimilation [9,76,77,78]. All other cyanobacteria and microalgae examined possess a standard form of the PCRC, though there are several phylogenetic variations in the regulation of the pathway (see [7] for a recent review).

6.2. Macroalgae and Seagrasses

According to early works, it was claimed that the common macroalga Ulva must be a C₄ alga based on its low Ci-compensation point and O₂-insensitive photosynthesis, as well as its high activities of PEPC and PEPCK relative to Rubisco. However, that was decided before ‘biophysical’ CCMs became known and accepted, and marine C₄ macroalgae were later seen as rare exceptions (e.g. we don’t know of any except Udotea flabellum [79]. Rather, from the mid-70s, ¹⁴C-pulse/¹²C-chase experiments showed that several macroalgal species including Ulva lactuca [80], another Ulva sp. [24], and what was then called Enteromorpha (now Ulva) compressa [82] featured a typical C₃ incorporation pattern with phosphoglycerate (a 3-carbon compound, thus C₃-metabolism plants) formed as the first compound in the PCRC cycle (Equation 2), they were all C₃ algae. Thus, it may be that also other macroalgae previously considered as C₄ are C₃ with additional ‘biophysical’ CCMs. While some workers found high levels of PEPC and PEPCK in Ulva prolifera [83], they still agree that this species is basically of the C₃ type. Another caution: the genus Ulva has at least 100 separate species [84], but they are often hard to tell apart. In our experience, however, all those Ulva forms (or alleged species) we have investigated perform similarly in terms of photosynthetic characteristics, including the feature of switching between external CA-mediated and the direct-HCO₃^- uptake mode. Low-level CAM has also occasionally been reported for members of the Fucales (see review by [85]); however, that too is a very rare exception.

Some macroalgae can live with diffusional CO₂ acquisition only. These include some red algae [62,63] growing in low-light and low-temperature environments where, apparently, photosynthetic rates are so low as not to require a CCM. On the other hand, this would also alleviate the need for energy to drive a CCM.

We conclude from parts 5 and 6 that marine photosynthesisers in general possess CCMs that are different from the ‘biochemical’ ones of terrestrial C₄ and CAM plants and, rather, are ‘biophysical’ as based on HCO₃^- utilisation.

7. Inorganic Carbon Acquisition in Various Marine Environments

Active acquisition of Ci via CCM activity is downregulated by elevated CO₂ supply where CO₂ supply by diffusion is sufficient to meet the needs of growth. CCMs appear also to be downregulated under low light. This seems to be the case for a number of red macroalgae grown at low light [86] as well as for the freshwater cyanobacterium Anabaena variabilis [87] where light supply is insufficient to drive CCM activity. Energy constraints on CCM activity are also seen under conditions of P-limitation (see [88] and references therein). In the marine diatom Chaetoceros muelleri, external CA and CCMs were up-regulated as cell density in cultures increased to a point where uncatalyzed rates of CO₂ supply from HCO₃^- were insufficient to meet the demands of carbon fixation [15]. It has been suggested that CCMs allow increased nutrient use efficiencies under nutrient limiting conditions, but evidence for N-limitation and Fe-limitation is mixed and a detailed discussion of environmental effects on carbon acquisition and CCM regulation is provided by [88].

Regarding macrophytes, we can give three examples of how photosynthetic Ci-acquisition modes of the prolific algae Ulva spp. can influence other macrophytes. Firstly, the unique ability of Ulva spp. to switch between external-CA-mediated HCO₃^- use and the very efficient direct uptake of the ion (making the alga 10 times more efficient in HCO₃^- use than when in the external-CA mode, see Section 6) can be connected with the climates where they grow. For example, Ulva fasciata growing in the Eastern Mediterranean where conditions of high temperatures and high irradiances are conducive to high photosynthetic rates, its HCO₃^- exchange with OH^- [22] will generate high pH values in their DBL and in isolated surroundings, causing CO₂ concentrations to approach zero. If so, epiphytic or other algae will be unable to grow in its surroundings. However, when this Ulva is transferred to the temperate, lower irradiance, waters of the Swedish West Coast, it switches to the external-CA mode [58]. Secondly, Ulva intestinalis can generate its own high-pH surroundings by HCO₃^-/OH^- exchange in isolated rockpools where the conditions during summertime are also conducive to high photosynthetic rates. Thus, the photosynthetic mode of this alga can effectively hinder other genera from the nearby open waters, which feature external-CA mediated Ci acquisition, to grow there [89]. Thirdly, it was shown that the generally higher photosynthetic rates of Ulva lactuca than of the seagrass Zostera marina could increase the pH and, so, lower the Ci and CO₂ and increase the O₂ concentrations to such values that they caused the seagrass to photorespire (71). Similarly, it was shown that ulvoid algae could deplete seawater conditions such that it affected the subtropical seagrass Thalassia hemprichii [90]. We estimate that future research will bring about more of these interactions based on photosynthetic capabilities.

8. Future Scenarios

Our planet is currently experiencing a period of rapid and intense environmental change which will result in elevated sea surface temperatures, higher CO₂ concentrations, lower ocean pH (ocean acidification, OA), and increased nutrient limitations due to enhanced stratification of the water column. All of these factors are likely to affect the physiological performance and population compositions of aquatic phototrophs. Consequences of these climate-induced changes have recently been reviewed [91].

At first sight, it would be predicted that aquatic photoautotrophs without a CCM, and thus showing a low affinity for CO₂ or HCO₃ ⁻ use in Ci acquisition, might show improved performance in the higher CO₂ environments of the future [92,93]. That said, such predictions have not always been borne out, at least in macroalgae [94] analysed reports on OA effects on a range of micro- and macroalgae and showed that in most species there was little or no influence of OA on performance, though in other cases there was a range of effects from decreases in photosynthesis and growth to large stimulatory effects. In the study of macroalgae by [95], of 55 macroalgae possessing CCMs, 21 showed no response to elevated CO₂, 15 exhibited increased growth and five had decreased growth relative to that in ambient seawater. Of the only five non-CCM species that were tested, three showed unaffected growth rates but enhanced growth in the other two species. However, it is clear that effects on growth are highly variable across taxa and not always predictable from the relationship between photosynthesis and Ci.

9. Summary

• Marine photoautotrophs in general use the 120 times higher HCO₃^- than CO₂ concentration in seawater for their photosynthetic needs;
• There are several ways in which cyanobacteria and microalgae can acquire Ci from seawater, including diffusion or active transport of CO₂. For many microalgae, however, as well as for macroalgae and seagrasses, the most common way is to convert HCO₃^- to CO₂ via membrane-bound CA activity associated with the periplasmic space. Another, more efficient way to acquire HCO₃^-, is by its direct uptake, mediated, at least in the macroalga Ulva, by an anion exchange protein bound to the plasma membrane;
• Because marine photosynthesisers contain Rubiscos with lower affinities for CO₂ than terrestrial C₃ plants, and given the slow diffusional supply of this Ci form in seawater, they are in need of (and typically possess) CCMs in order to partly or fully (depending on species) saturate Rubisco with CO₂ so as to optimise photosynthetic- and growth rates. Some algae, however, can under low irradiance utilise only CO₂ by diffusion;
• The ‘biophysical’ CCMs of marine photoautotrophs are different from the ‘biochemical’ CCMs of terrestrial C₄ and CAM plants as they rely on extracellular HCO₃^- supplying CO₂ to their Rubiscos;
• Photoautotrophs using C₄ and CAM pathways for inorganic carbon fixation are very rare in marine environments, but C₄ metabolism may in some cases have an anaplerotic carboxylation role;
• While many macroalgae and all seagrasses investigated in laboratory conditions require additional CO₂ to fully saturate carbon fixation, their performance in situ may be different such that they are closer to CO₂ saturation without additional CO₂ or Ci additions;
• Responses to future changes in CO₂ levels would appear to be very species dependent and also influenced by modulation of CCM activity by other environmental conditions such as light and nutrient levels.