Responses of Freshwater Calcifiers to Carbon-Dioxide-Induced Acidification

Round 1

Reviewer 1 Report

1. in Materials and Methods section, I would like to suggest a schematic diagram of graph showing the setup of the experiments, which will be more convenient for readers

2. I recommend using the unit vvm, volume gas per volume water per minute, during gas input. And how would the equilibrium be evaluated?

3. line 31: compressed CO2, do you mean pure CO2, or mixture with air?

4. Different species might show quite different range of tolerance to CO2 concentrations. The response of pea clam was complex, but the elucidation should be more comprehensive instead of being only descriptive.

4. The calcification process in most living organisms is in a dynamic pattern, the synthesis and dissolution, which occurred at the same time. I understand that the experiment has been conducted in a long period of observation. But, it would be nice if there is a control experiment with only the shells instead of the whole living organisms, ruling out the natural physicochemical effects on calcified shells.

5. pCO2 in the atmospheric and in culture medium are two different values. How author explain the consistency? The CO2 concentration in air and the resulting CO2 in the water (also the concentration of HCO3-, H2CO3).

Author Response

Thank you very much for your time reviewing this manuscript. Please see below for our responses to your comments.

1. in Materials and Methods section, I would like to suggest a schematic diagram of graph showing the setup of the experiments, which will be more convenient for readers

Comment 1: This reviewer suggested that a schematic diagram be included showing how the experiment was setup. We agree and have included this as Figure 1.

2. I recommend using the unit vvm, volume gas per volume water per minute, during gas input. And how would the equilibrium be evaluated?
3. line 31: compressed CO2, do you mean pure CO2, or mixture with air?

 

5. pCO2 in the atmospheric and in culture medium are two different values. How author explain the consistency? The CO2 concentration in air and the resulting CO2 in the water (also the concentration of HCO3-, H2CO3).

Comments 2, 3, and 5: This reviewer also suggested clarifying the CO₂ dosing methodology and the maintenance of CO₂ disequilibrium between the atmosphere and the water. This included stating the volume of gas per volume of water per minute added and whether we added pure or an air-mixture of CO₂. We thank the reviewer for pointing out that we neglected to mention the concentration of CO₂ added. It was pure CO₂ and we have added this description to several places in the manuscript. As for the dosing rate relative to the water dosing rate, the pH stat controller added either pure CO₂ or ambient air as needed to maintain the CO₂ treatments. It was not set to a constant rate, and we have added text to clarify this point. The disequilibrium between the treatment and the atmosphere was maintained with the pH stat controller. We assisted this process by covering the sumps with a plastic sheet and covering the replicate tanks with plastic tops to isolate the treatment and sump waters from the room environment.

4. Different species might show quite different range of tolerance to CO2 concentrations. The response of pea clam was complex, but the elucidation should be more comprehensive instead of being only descriptive.
5. The calcification process in most living organisms is in a dynamic pattern, the synthesis and dissolution, which occurred at the same time. I understand that the experiment has been conducted in a long period of observation. But, it would be nice if there is a control experiment with only the shells instead of the whole living organisms, ruling out the natural physicochemical effects on calcified shells.

Comment 4a: We thank the reviewer for challenging us to better elucidate the response of the pea clam. We had included a brief description of potential mechanisms driving its response but have expanded the discussion to better explore why it responded differently, specifically the unique life history of the pea clam.

Comment 4b: Lastly this reviewer has pointed out the complexity of attributing patterns of net calcification to the potentially independent processes of gross calcification and gross dissolution during the experiment. We wholeheartedlly agree with this and are also fascinated by these questions. However, this approach would require independent assessment of gross dissolution rates, which was beyond the the scope of the present study – although it is certainly something worthy of exploring in a future experiment.

Reviewer 2 Report

 

This manuscript studies the effect of water acidification on 3 species of mollusks and one crustacean. The main conclusions of the authors is that water acidification leads to a decrease in the calcification rate. This paper contains several flaws:

- The first one is in the methodology: the calcification rate was calculated as the fractional change in shell mass, assuming a density of 2.71 g/cm3. This is the density of calcite, however Margaritifera margaritifera has a shell made of both calcite and aragonite, and the density of aragonite is 2.93 g/cm3. Corbicula fluminea has a shell composed entirely of aragonite. This means that the fractional change in shell mass is incorrect and has to be recalculated. It is not clear to me how the fractional change in shell mass was calculated for the crayfish – how can the authors be sure that changes in weight are not due to growth? Did the authors take into account the moulting cycle, when the CaCO3 that reinforces the carapace is dissolved and stored as amorphous calcium carbonate?

- Only 3 individuals per tank were measured, over only 3 conditions of pCO2. I find this number hardly enough for convincing statistical analysis. Looking at the pearlshell mussel, for example, there are two points that maintained more or less the same growth rate over the different pCO2 conditions. So is the downward trend that was plotted real? Do the 3 pCO2 conditions for the pea clam justify drawing a parabolic trend? I am not convinced of this.

- I am surprised that the authors did not look at the shell structure to see if there were differences with the pCO2 – thickness of the mineral, morphology of the crystals, etc. This would have given stronger information on the effect of acidification.

Author Response

We thank the reviewer for their time in improving our manuscript. Please see below for our responses to your comments.

1. The first one is in the methodology: the calcification rate was calculated as the fractional change in shell mass, assuming a density of 2.71 g/cm3. This is the density of calcite, however Margaritifera margaritifera has a shell made of both calcite and aragonite, and the density of aragonite is 2.93 g/cm3. Corbicula fluminea has a shell composed entirely of aragonite. This means that the fractional change in shell mass is incorrect and has to be recalculated. It is not clear to me how the fractional change in shell mass was calculated for the crayfish – how can the authors be sure that changes in weight are not due to growth? Did the authors take into account the moulting cycle, when the CaCO3 that reinforces the carapace is dissolved and stored as amorphous calcium carbonate?

We are grateful to the reviewer for pointing out our oversight in assuming all shell material here had a density of 2.71. Using our buoyant and air weights of the pure shell material, we have estimated that the shell of the mussel, Margaritifera margaritifera, has a density of 2.707 and the shell of the clam, Corbicula fluminea has a density of 2.811, consistent with a larger contribution of aragonite in the clam shell. We have updated our growth rate calculations and resulting models accordingly.  For the crayfish and their molting cycle, we had already excluded crayfish that molted immediately prior to weighing and we have added this detail to the text. We continue using a density of 2.71 g cm^-3 for the crayfish shells.  

2. Only 3 individuals per tank were measured, over only 3 conditions of pCO2. I find this number hardly enough for convincing statistical analysis. Looking at the pearlshell mussel, for example, there are two points that maintained more or less the same growth rate over the different pCO2 conditions. So is the downward trend that was plotted real? Do the 3 pCO2 conditions for the pea clam justify drawing a parabolic trend? I am not convinced of this.

We thank the reviewer for also pointing out their concerns over our statistical analysis, which we believe is in reference to our miscommunication over the former Figure 3, now Figure 4, and the appearance that we only measured three individuals per CO₂ treatment, for a total of 9 data points. However, rather than 3 treatments with 3 individuals each, we had 3 treatments with three replicate tanks per treatment, for a total of nine total tanks. Furthermore, each replicate tank had multiple individuals; Figure 4 shows just the tank averages, not the total number of individuals. Therefore, trends shown are for many more than 9 individuals per species, and totaled 33 crayfish, 95 Asian clams, 62 pearlshell mussels, and 85 pea clams. We have clarified this issue in the figure legend and also included the total numbers of individuals per treatment in Table 2.  The large number of individuals employed in the pea clam experiment (85, versus the presumed 9) should also alleviate any concerns about too few individuals being assessed to identify the parabolic shape of the relationship between pea clam growth and pCO2. We apologize for the lack of clarity on this matter and thank the reviewer for bringing this to our attention.

3. I am surprised that the authors did not look at the shell structure to see if there were differences with the pCO2 – thickness of the mineral, morphology of the crystals, etc. This would have given stronger information on the effect of acidification.

Lastly, this reviewer also suggested that more information regarding shell properties (thickness, crystal morphology, etc) would provide more information on the effect of acidification on these freshwater calcifiers. Although we agree that this line of inquiry would be a valuable contribution to understanding how these species will fare in the future, we believe that this topic falls outside the scope of these particular experiments, but agree that this is something worth pursuing in a subsequent study.

Reviewer 3 Report

 

This manuscript investigates the impact of elevated pCO2 on freshwater calcifiers and the impact of the associated rising in pH on calcium carbonate shell formation. Four representative species were exposed to low, moderately elevated, and extremely elevated pCO2 conditions and cultivated in long-term experiments.

 

The manuscript is beautifully written, short and precise and gives great insight into the effect of rising pH on freshwater calcifiers – an important topic of concern at the present. Such knowledge currently mostly exists for marine calcifiers, therefor the data collected can be considered a valuable scientific contribution. The manuscript is therefore recommended for publication in J. Mar. Sci. Eng. with very few suggestions for improvement. The manuscript form and language quality is excellent and needs no further improvement.

 

My only main concerns is the quality of the correlations: As the authors only investigated 3 different sets of pCO2, these correlations (3x linear, 1x parabolic) to me are not convincing. For example, the relative growth rate of the Asian clam or the pearlshell mussel could also easily be fitted by a parabolic function. This is also reflected in the very minimal statistic differences between these fits in Table 1. An additional dataset at around 2000-2500 µatm would have been very helpful to improve the dataset and thereby get a more convincing fitting.

 

The same is true for the feeding experiments, which simply could have had more samples to improve statistical significance. In my opinion, these issues should be addressed, if further experiments are not feasible (as they are probably time-intensive). An option could be to refrain from the fitting (as done for the feeding experiments) and discuss it more qualitatively.

 

 

Author Response

We thank the reviewer for their kind words regarding the writing quality of the manuscript. We very much appreciate that the effort we put into preparing the manuscript was noticed and well received. Please see below for our response to your comments. 

1. My only main concerns is the quality of the correlations: As the authors only investigated 3 different sets of pCO2, these correlations (3x linear, 1x parabolic) to me are not convincing. For example, the relative growth rate of the Asian clam or the pearlshell mussel could also easily be fitted by a parabolic function. This is also reflected in the very minimal statistic differences between these fits in Table 1. An additional dataset at around 2000-2500 µatm would have been very helpful to improve the dataset and thereby get a more convincing fitting.

The same is true for the feeding experiments, which simply could have had more samples to improve statistical significance. In my opinion, these issues should be addressed, if further experiments are not feasible (as they are probably time-intensive). An option could be to refrain from the fitting (as done for the feeding experiments) and discuss it more qualitatively.

The reviewer’s concerns center around the quality of the correlations between growth rates and pCO₂ and, to some extent, feeding rates and pCO₂. Specifically, there is a larger gap between the intermediate and high pCO₂ treatments than between the intermediate and low pCO₂ treatments. Qualitatively, this makes an exponential or parabolic model for the Asian clam and pearlshell mussel appear to be equally good at describing the data. Although we agree that additional data in the 2000-2500 uatm range would help inform our identification of the best-fit models, obtaining additional data at these intermediary pCO2 conditions is not possible because the experiment is no longer running and neither of the co-authors are at the field station where the work was conducted. We are also hesitant to abandon these models for more qualitative descriptions or even categorical statistics, like an ANOVA approach, because we  feel that the regression approach better accounts for the variability in pCO₂ conditions between tanks even though it may favor simpler models (linear over parabolic or exponential), especially given that the replicate tanks had slightly different chemistries and thus provide additional inferential power by spanning more of the experimental domain—which is only realized through the regression approach. We have added text to the discussion addressing this matter. For the feeding experiments, we agree that more samples would have been ideal (along with feeding choice experiments and feeding rates of the other species, etc.). Although an experiment can almost always be improved by making additional observations, we were limited in the time and resources that we could allocate to this project. Furthermore,  the project has already been completed and it is, therefore, not feasible to obtain any additional data for these feeding experiments. That said, we believe that there is inferential power in the feeding data that were obtained, and we believe that they enhance the manuscript as presently included.

Round 2

Reviewer 1 Report

Most of my concerns have been addressed.

I would like to recommend acceptance.