Dear Authors,

I very much enjoyed reading this paper and believe it is a fantastic contribution to Biogeosciences. I have left very minor comments which need addressing/acknowledging, but other than that, I look forward to the manuscript being published.

All the best,

Adam Woodhouse (signed review)

General comments:

The manuscript entitled “The impact of essential climate variables on respiration rates in subpolar and polar planktonic foraminifera” by Armitage et al. reported the relationship between environmental parameters and respiration rates of polar and subpolar planktonic foraminifera species. Temperature effects were discussed in detail, and the authors showed that Neogloboquadrina pachyderma, a polar species that is often utilized for paleoenvironment reconstruction, has relatively stable respiration rates over a wide range of temperatures with low Q₁₀. This finding alleviates our concerns on potential respiration effect on foraminiferal test geochemistry with regard to N. pachyderma. They also conducted micro-Xray scanning to calculate biovolume more precisely, which allowed discussion on allometric scaling of respiration for generalization.

This study is important to gain our understanding of the basic metabolic activity of foraminifera under different temperature conditions, as well as ground the validity of species to be used in paleoenvironmental reconstruction. The manuscript is overall well-written, with detailed methods used and carefully discussed. However, I have several major concerns regarding the statistical treatment of the data and, in particular, the interpretation of the results. In several places, the analyses rely on limited datasets or assumptions that are not fully justified, and some conclusions appear to extend beyond what can be robustly supported by the data. I believe that addressing the points raised below—especially by reconsidering the statistical approaches and tempering some of the broader interpretations—would substantially strengthen the manuscript.

Specific comments:

1. Statistical analysis on respiration rates and other parameters

The relationships between respiration rate and essential climate variables (ECVs) are evaluated primarily through separate pairwise correlation analyses (reporting r² and p values for each parameter, Table 3). While this approach may be useful as an exploratory analysis, it has important limitations that should be acknowledged. Many of the environmental variables considered (e.g., temperature, nutrients, salinity, DIC) are likely to be intercorrelated due to shared environmental gradients, such as water mass structure or seasonality. As a result, the reported correlations do not allow the independent effects of individual parameters on respiration to be disentangled. In addition, testing multiple environmental variables separately raises concerns about multiple comparisons, which may inflate the likelihood of detecting spurious significant relationships. I recommend either applying a multivariate framework (e.g., multiple regression or related approaches).

As I explain in the next part, correlation analysis for N. incompta needs to be reconsidered, since the datasets (based on 3 stations) cover a narrow range of each variable.

2.  Q10 calculation for N. incompta

For this species, respiration rates were measured at only three temperatures (10, 13, and 14 °C, in situ), covering a very narrow temperature range. In addition, the respiration rates recorded the highest at 13°C, and declined at 14°C. The calculation of Q10 based on the present data set appears problematic. Q10 assumes a monotonic, approximately exponential increase in metabolic rate over a sufficiently wide temperature range, under conditions where temperature is the primary limiting factor. Moreover, respiration peaked at 13 °C and declined at 14 °C, indicating a non-monotonic response and suggesting that the measurements may already span an optimal temperature or the onset of thermal stress. Under these conditions, the fundamental assumptions underlying Q10 are not met, and the resulting values are difficult to interpret physiologically. I therefore suggest either refraining from calculating Q10 or clearly stating that any estimated Q10 values are highly tentative and limited to a restricted temperature interval.

3. Difference between spinose and non-spinose

The authors discuss the difference in Q₁₀ values between N. pachyderma and T. quinqueloba relating the morphology and trophic mode of the species. It is true that T. quinqueloba is a spinose species, but this species is a “short-spined” species that has completely different ecology and physiology from typical spinose-species like Globigerinoides, Globigerina, Globigerinella, Orbulina, etc. Specifically, T. quinqueloba is not a carnivorous species, nor a symbiont-bearing species, nor an oligotrophic-adapted species. Presence of spines is an adaptation for planktonic lifestyle, but since the non-spinose Neogloboquadrina species also share the shallow habitat as is presented in the sample metadata (Table 1), morphological difference (presence or absence of spines) is not meaningful to explain the Q₁₀ difference, I would say. As the authors noted at L443–445, it is true that non-spinose N. incompta showed relatively high Q10 (although it needs reconsideration as I pointed out above), which already collapses the validity of spinose/non-spinose comparison. I would say it’s just species-specific.

Anyway, this part (L431–445) needs to be reconsidered.

4. Generalization of metabolic allometry

Representing the biovolume–respiration scaling relationship of planktonic foraminifera with data from other publications is interesting and potentially valuable. However, I don’t fully understand the discussion on “crossover point” in L463–471. What exactly does the crossover point in Fig. 8 mean?

Moreover, I am concerned that the interpretation of the resulting scaling exponent may be overstated. The authors’ statement that foraminiferal metabolism is somewhat “intermediate” between protists and metazoans more complex metazoans relies on cross-study comparisons that involve heterogeneous data sets, differing methodologies, and taxonomically broad groups. Given these uncertainties, the observed position of the foraminiferal scaling exponent relative to other organisms may reflect dataset composition or methodological differences rather than fundamental differences. I therefore suggest toning down this interpretation and framing it more explicitly as a hypothesis or conceptual possibility.

5. T. quinqueloba symbiotic ecology

In Hemleben et al. (1989), it is indeed written that T. quinqueloba possesses symbionts, but no data are presented. Stangeew (2001) interpreted this species as symbiotic, based on the statement in Hemleben et al. (1989), which also does not show any evidence for the presence of symbionts on this species. Takagi et al. (2019) classified T. quinqueloba as a non-symbiotic species based on the absence of active chlorophyll fluorescence (photosynthetic activity). In this sense, “…their presence remains elusive (e.g., Takagi et al. 2019, ….)” is not appropriate. As far as I know, no positive data/evidence of the presence of symbionts for T. quinqueloba is available. Since the authors’ observation is also in alignment with the absence of symbionts for T. quinqueloba, I think it’s safe to say the specimens they used were non-symbiotic.  

6. Biovolume and test size

It is usually the case that the final chamber of collected foraminifera specimens is empty. In that case, biovolume estimation from the whole test would cause overestimation, since the final chamber generally has the largest volume. In this study, was this point considered? Since the experiments were conducted at different time points from collection (within 24 hrs without food supply for CE23011, and fed specimens within 11 days for 2024 samples), specimens conditions might have been different. Ideally, filled or not needs to be checked, and the biovolume needs to be corrected by excluding the empty chambers. If this is not possible, at least, please make remarks on the cytoplasm volume, that it is not always equal to the cavity volume. In Burke et al. (2025), 75% of cavity volume was applied as biovolume. This is an alternative way to take into account the void part.

Minor points

L313: Hemleben (1989): Hemleben et al. (1989)

L313: Stangeew et al. (2001): Stangeew (2001)

Fig.2  The illustration of the Unisense logos are confusing. Since it resembles to planktonic forams (maybe the logo derives from forams), I thought, at first glance, the specimens are located in those boxes. Please delete the logo. In addition, the cable of the “calibration chamber” is not connected anywhere. Is this correct?

Fig. 5  Why is the y-axis for panel (a) (Temperature) alone on a log scale whereas the others are in linear scale?