Contrasting calcification responses to ocean acidification between two reef foraminifers harboring different algal symbionts



[1] Ocean acidification, which like global warming is an outcome of anthropogenic CO2 emissions, severely impacts marine calcifying organisms, especially those living in coral reef ecosystems. However, knowledge about the responses of reef calcifiers to ocean acidification is quite limited, although coral responses are known to be generally negative. In a culture experiment with two algal symbiont-bearing, reef-dwelling foraminifers, Amphisorus kudakajimensis and Calcarina gaudichaudii, in seawater under five different pCO2 conditions, 245, 375, 588, 763 and 907 μatm, maintained with a precise pCO2-controlling technique, net calcification of A. kudakajimensis was reduced under higher pCO2, whereas calcification of C. gaudichaudii generally increased with increased pCO2. In another culture experiment conducted in seawater in which bicarbonate ion concentrations were varied under a constant carbonate ion concentration, calcification was not significantly different between treatments in Amphisorus hemprichii, a species closely related to A. kudakajimensis, or in C. gaudichaudii. From these results, we concluded that carbonate ion and CO2 were the carbonate species that most affected growth of Amphisorus and Calcarina, respectively. The opposite responses of these two foraminifer genera probably reflect different sensitivities to these carbonate species, which may be due to their different symbiotic algae.

1. Introduction

[2] Nearly one-third of all anthropogenic CO2 produced since the beginning of industrialization has been absorbed by the ocean [Sabine et al., 2004; Sabine and Feely, 2007]. This has caused a decrease in seawater pH of 0.1 [Raven et al., 2005] and altered the carbonate chemistry of surface seawater. Together, these changes are referred to as ocean acidification. By the end of 21st century, the partial pressure of atmospheric CO2 (pCO2) is predicted to reach 500–1000 μatm under various IPCC SRES scenarios [Solomon et al., 2007]. These changes are expected to decrease seawater pH by an additional 0.3–0.4 units and the carbonate ion concentration by 50% [Orr et al., 2005]. Such alteration of seawater chemistry can influence the calcification of marine calcifiers. Various studies have revealed potentially dramatic responses in a variety of calcareous organisms to the range of pCO2 values projected to occur over this century. For example, an 8–14% reduction of planktonic foraminifer calcification and a 25% decrease of molluscan calcification [Bijma et al., 1999; Spero et al., 1997; Gazeau et al., 2007]. The sensitivity of the response varies both among and within species, and some taxa of coccolithophores and sea urchins even show enhanced calcification in environments with higher pCO2 [Iglesias-Rodriguez et al., 2008; Doney et al., 2009; Ries et al., 2009]. In particular, different populations of Emiliania huxleyi have shown decreased, increased, or unchanged calcification in response to higher pCO2 [Fabry, 2008].

[3] According to Hoegh-Guldberg et al. [2007], aragonite saturation will drop below the threshold for major changes to coral communities within this century, which will result in less diverse reef communities and carbonate reef structures that fail to be maintained. Carbonate mineral production in coral reefs is largely supported by the calcification of corals, coralline algae, and large benthic foraminifers. Among these, reef-dwelling foraminifers and coralline algae mainly produce carbonate shells composed of high-Mg calcite, which has generally higher solubility in seawater than low-Mg calcite or aragonite [Morse et al., 2006]. The empty tests of these organisms, released by reproduction or death, are entrained at the reef crest by waves and transported by currents, and contribute to the formation and retention of sand beaches and coastal landforms [Hohenegger, 2006; Fujita et al., 2009].

[4] Reef-calcifying organisms other than corals are also both biogeochemically and ecologically important, but they have been much less studied to date [Kuffner et al., 2008; Doney et al., 2009]. Kuroyanagi et al. [2009] cultured the reef-dwelling benthic foraminifer Marginopora (Amphisorus) kudakajimensis in pH-controlled seawater and showed that lower pH decreased both the shell weight and diameter. However, in their experiment they controlled the seawater pH by adding a strong acid or base, which also alters the alkalinity and does not precisely reproduce ocean acidification conditions.

[5] During ocean acidification, carbonate ion (CO32−) in seawater is decreased and at the same time bicarbonate ion (HCO3) and CO2 are increased. The former change may negatively influence the calcification of foraminifers by reducing the calcium carbonate saturation state (Ω) of the seawater, whereas the latter changes may have a positive effect through enhancement of symbiont photosynthesis. Thus, ocean acidification may affect net calcification of foraminifers in either direction. Sensitivity differences to each of these changes may determine the overall response of a particular species to ocean acidification. Therefore, to accurately estimate the impacts of ocean acidification, the effect of each change should be examined separately.

[6] In this study, we focused on a varying responses to ocean acidification between species of algal symbiont-bearing, reef-dwelling foraminifers by conducting a series of culture experiments. We used a high-precision pCO2 control system to evaluate the effects of ongoing ocean acidification on foraminiferal calcification under possible near-future pCO2 conditions. To evaluate the impact of HCO3 and CO32− concentration changes in seawater on net calcification of foraminifers separately, we also conducted a culture experiment in which seawater was chemically manipulated to vary the HCO3 concentration under a constant CO32− concentration.

2. Materials and Methods

[7] We selected two genera (three species) of large, algal symbiont-bearing benthic foraminifers commonly found on coral reefs in the northwest Pacific. Amphisorus kudakajimensis and the closely related species Amphisorus hemprichii have porcelaneous shells and harbor dinoflagellates as their symbiont algae [Lee, 1998]. Typically, these species are found on macroalgae in the reef moat [Hohenegger, 1994]. Calcarina gaudichaudii has a hyaline shell and is host to diatom endosymbionts [Lee, 1998]; it typically lives on algal turf on the reef crest [Hohenegger, 1994].

[8] Mature living foraminifer individuals were collected from Okinawa, Japan, in early May 2010; A. kudakajimensis and A. hemprichii were collected from Aka Island (26°39′N, 127°51′E) and C. gaudichaudii from Ikei Island (26°39′N, 127°99′E). In the laboratory, the mature individuals reproduced asexually, each producing 500–1000 clonal (i.e., genetically identical) individuals. We cultured three separate clone populations of each species. By using clonal individuals for culture experiments, genetic influences can be eliminated. Moreover, since the initial size and weight of clone individuals are practically identical, weight differences of clone populations between treatments can be assumed to directly reflect differences in calcification during the experimental period. Therefore in the present paper we focus on the variation of shell weight, not shell size, between treatments.

2.1. Ocean Acidification (OA) Experiment

[9] We cultured A. kudakajimensis and C. gaudichaudii individuals in seawater under five different pCO2 conditions. Highly precise and stable pCO2 conditions (pCO2 levels were maintained mostly within 5% throughout the experimental period) were achieved by using a high-precision pCO2 control system called the AICAL system [Fujita et al., 2011]. In this system, filtered seawater (pore size 1 μm) is exposed to a gas mixture of CO2 and dilution air. pCO2 of seawater flowing out from a bubbling tank was measured directly and maintained at the desired level by continuously regulating the ratio of CO2 in the gas mixture. The five treatment levels used, 245, 375, 588, 763, and 907 μatm, represented pre-industrial (Low pCO2), present-day (Control), and three near-future (High pCO2 1–3) pCO2 conditions, respectively (Table 1 and Figure 1).

Figure 1.

(a) Bicarbonate and (b) carbonate ion concentrations, (c) pCO2, and (d) pH in the experimental seawater of each treatment in the ocean acidification (closed blue circles) and constant carbonate (open red squares) experiments.

Table 1. Carbonate Chemistry Speciation for Each Treatment of the Ocean Acidification (OA) and Constant Carbonate (CC) Experimentsa
TreatmentpH at 25°CTADICpCO2HCO3CO32−CO2ΩcalΩarg
  • a

    Total alkalinity (TA) and pCO2 data in the OA experiment and TA and pH (total hydrogen ion scale) in the CC experiment are the means over the experimental period. Other values were calculated from these values using CO2SYS software [Pierrot et al., 2006], the temperature and salinity given in the table, and the apparent dissociation constants for carbonic acid of Mehrbach et al. [1973], refit by Dickson and Millero [1987]. Units of concentration are μmol kg−1 and those of pCO2 are μatm.

Ocean Acidification Experiment (∼27.1°C, Salinity 34.1)
Low pCO28.23222241821245153627876.84.5
High pCO2 - 17.924222420025881823163164.02.6
High pCO2 - 27.826222420477631891135213.32.2
High pCO2 - 37.761222420759071932119242.91.9
Constant Carbonate Experiment (∼26.9°C, Salinity 34.5)
Low HCO38.081216518553541630216105.23.5
High HCO3 - 18.031237120714481841217125.33.5
High HCO3 - 28.014244721514861920218135.33.5
High HCO3 - 38.003249121975111966217145.33.5

[10] Ten to fifteen individuals of each clone population were sealed in an acrylic pipe cage constructed from 180 μm mesh nylon sheets and submerged in a 12-L aquarium filled with pCO2-controlled seawater. For each pCO2 treatment, two aquariums were prepared so that reproducibility could be assessed. Seawater was continuously supplied to the aquariums at the rate of 150 mL per minute. Cultured individuals were maintained for about 4 weeks in an indoor flow-through system at the same constant water temperature and light intensity under a 12 h:12 h light:dark cycle. They were not fed during the experimental period. After the experiment, foraminiferal shells were dried and their weights were measured separately using a Thermo Cahn C-35 microbalance, which can measure weights down to 0.1 μg with a precision (reproducibility) of less than 1.0 μg. The contribution of organic matter to the dry weight in this procedure is less than 3% [Fujita and Fujimura, 2008] with negligible differences among treatment conditions.

2.2. Constant Carbonate (CC) Experiment

[11] To separately evaluate the impact on calcification of HCO3 and CO32− concentration changes in seawater, we conducted a culture experiment with A. hemprichii and C. gaudichaudii using seawater chemically manipulated to vary the HCO3 concentration while maintaining a constant CO32− concentration. The seawater carbonate chemistry was manipulated by the following two steps. First, we altered the total alkalinity (TA) of the seawater by adding Na2CO3 or HCl, and then we manipulated the dissolved inorganic carbon (DIC) content by bubbling either CO2 gas, air, or CO2-free air, selected as appropriate for the required direction of DIC change. The seawater chemistry was modified just before the seawater was to be used in a culture experiment. The carbonate chemistry in the five treatments in this experiment was designed to represent a similar range of HCO3 concentrations to those in the OA experiment while maintaining a constant CO32− concentration similar to that of the control treatment in the OA experiment (Table 1 and Figure 1).

[12] About 10 individuals of a clonal population were sealed in a 120-mL glass vial filled with chemically manipulated seawater. We prepared duplicate sets of three clone populations for each species. The culturing seawater was replaced every week, and the glass vials were tightly sealed with minimum headspace to prevent any CO2 gas exchange between the seawater and the atmosphere. The duration of cultivation was 4 weeks for A. hemprichii and 5 weeks for C. gaudichaudii.

2.3. Statistical Analysis

[13] A preliminary graphical plot of measured shell weight showed a long-tailed distribution, suggesting that shell length rather than weight is normally distributed. Therefore, before the statistical analysis, we calculated the square root of the measured shell weights for Amphisorus spp. and the cube root for C. gaudichaudii, in accordance with the species-specific geometric direction of skeletal growth (Amphisorus spp. have disc-shaped shells, and Calcarina shells are spherical with spikes), which results in their having different surface-to-volume ratios [Irie and Adams, 2007; Kuroyanagi et al., 2009]. The transformed shell weight was analyzed by ANOVA in which pCO2, clone, replicate tank nested within pCO2, and all possible interactions were considered fixed effects. ANOVA was followed by Tukey's HSD tests to find significant differences in the focal factors among levels (α = 0.05). We used JMP statistical software (version 7.0.1. SAS Institute Inc.) for all statistical analyses.

3. Results

3.1. OA Experiment

[14] The main effects of pCO2 on shell weight were independent of clone or rearing tank in both species (Tables S1 and S2 of the auxiliary material). Higher seawater pCO2 led to smaller mean shell weight in A. kudakajimensis (Figure 2a). Conversely, in C. gaudichaudii seawater pCO2 and shell weight were positively related (Figure 2b). Difference in clonal origin was also statistically significant, reflecting the differences in the initial weights of the clone populations (Tables S1 and S2 of the auxiliary material). In contrast, the significant effect observed between replicate tanks might reflect a difference in environmental conditions unrelated to pCO2. As temperature (monitored, not shown) and seawater composition other than carbonate chemistry were identical in all tanks, the difference might be attributable to insufficient randomization of light conditions, but we cannot ascertain the reason from the available data. However, we were able to reduce the influence of the difference on the result by combining the results of replicate tanks. It would also be possible to reduce this effect by increasing the sample size.

Figure 2.

Least mean square (± standard error) adjusted for the rearing tank of (a) the square root of the shell weight of A. kudakajimensis and (b) the cube root of the shell weight of C. gaudichaudii after the ocean acidification experiment. Arithmetic mean (± standard error) of (c) the square root of the shell weight of A. hemprichii and (d) the cube root of the shell weight of C. gaudichaudii after the constant carbonate experiment. The typical shapes of Amphisorus and Calcarina individuals are shown in Figures 2a and 2b, respectively.

3.2. CC Experiment

[15] No statistically significant trend in shell weight was found among the five treatments in either A. hemprichii or C. gaudichaudii (Figures 2c and 2d), but A. hemprichii exhibited a significant interaction between pCO2 and clone (Table S3 of the auxiliary material), suggesting that the reaction norm is variable among clones. However, Tukey's HSD test revealed no impact of pCO2 level on shell weight in two of the three clonal groups, and the pH effect was not unidirectional in the other group (growth rates in the Low and High-3 treatments were higher than in the other treatments). Similarly, neither the pCO2 × clone nor the tank × clone interaction was statistically negligible in C. gaudichaudii (Table S4 of the auxiliary material), but the Tukey's HSD test result suggested that shell weight was independent of pCO2 condition in five of the six tank × clone combinations (in the sixth combination, the test detected significantly slower growth in High-1 than in the other treatments).

4. Discussion

[16] A decreasing trend of A. kudakajimensis skeletal weight with lower pH was previously reported by Kuroyanagi et al. [2009], who used the acid addition method. We thus confirmed the negative impact of ocean acidification on this species by performing a precise perturbation experiment using the gas-bubbling method, which more realistically simulates ocean acidification.

[17] To shed light on the causes of the different calcification responses between Amphisorus (A. kudakajimensis and A. hemprichii) and Calcarina (C. gaudichaudii), we considered the combined results of the OA and CC experiments to evaluate which inorganic carbon species in seawater most affected calcification by foraminifers. As Amphisorus did not show any significant trend with higher pCO2 or lower pH in the CC experiment, it is highly likely that the CO32− concentration, and thus the saturation state of the seawater with respect to calcium carbonate (Ω), importantly influences calcification in Amphisorus.

[18] Despite lower CO32− and Ω, Calcarina showed an increase in net calcification with higher pCO2 in the OA experiment, but like Amphisorus, no significant trend in the CC experiment. The two experiments were designed to have a similar bicarbonate ion concentration range (Figure 1), so the upward trend in the OA experiment can probably be attributed to the increase in CO2, possibly through enhancement of symbiont photosynthesis, a phenomenon known as the CO2-fertilizing effect [e.g., Ries et al., 2009]. In the OA experiment pCO2 increased by as much as 140% compared with the control, whereas in the CC experiment pCO2 increased by only 30% (Table 1 and Figure 1). The different responses of Calcarina between the two experiments may have been due to this different pCO2 gradient. A positive calcification response to ocean acidification has also been reported in coccolithophores [Iglesias-Rodriguez et al., 2008], which also calcify and photosynthesize simultaneously.

[19] As one possible cause of these different sensitivities, we speculate that the type of symbiont influences the strength of the CO2-fertilizing effect. Calcarina hosts diatoms as its symbiotic algae, whereas Amphisorus hosts dinoflagellates. Both a single-species culture experiment [Wu et al., 2010] and a mesocosm bloom experiment [Engel et al., 2008] have shown that high-CO2 seawater is favorable to diatom growth. Moreover, Badger et al. [1998] pointed out that a rise in CO2 may lead to enhanced phytoplankton growth owing to the low affinity of the carboxylating enzyme (Rubisco) for CO2. Although it is difficult from our data to evaluate the importance of increased growth of the symbiont, it is possible that Calcarina acquires an increased amount of energy from its symbiotic diatoms under high pCO2 conditions, leading to enhanced calcification. On the other hand, Rost et al. [2006] reported that dinoflagellates use HCO3 as their carbon source, so their rate of carbon fixation may remain unaffected by fluctuating CO2 levels. Many laboratory studies of various coral species having dinoflagellates as their symbiotic algae have confirmed that coral calcification rates decrease with increasing pCO2 [Doney et al., 2009]; these results may indicate that the CO2-fertilizing effect of dinoflagellates is weak, or that the dependence of Amphisorus on photosynthesis is low [Lee et al., 1991].

[20] In conclusion, the results of our precisely pCO2-controlled culture experiment revealed that two genera of large benthic foraminifers, Amphisorus and Calcarina, showed contrasting net calcification responses, with Calcarina even showing enhanced calcification. Comparison of these results with those of a culture experiment under a constant CO32− concentration suggested that the negative response may be due to the decrease in CO32− in the seawater, and the positive response to the increase in CO2. We speculate that these different influences of seawater chemistry may be attributable to the different types of symbiotic algae hosted by Amphisorus and Calcarina.


[21] We are grateful to A. Iguchi, H. Kinjo, S. Ohki (University of the Ryukyus), M. Inoue, and K. Shinmen (the University of Tokyo) for laboratory assistance and valuable advice. This study was supported by the Acidification Impact on Calcifiers (AICAL) project funded by the Global Environment Research Fund A-0804 of the Ministry of the Environment of Japan and by a grant-in-aid from the Japan Society for the Promotion of Science to H. Kawahata (22224009). An anonymous reviewer and Jason Hall-Spencer provided helpful comments on the manuscript.

[22] The Editor thanks Jason Hall-Spencer and an anonymous reviewer for their assistance in evaluating this paper.