the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Jan 2025
Photosynthetic electron, carbon and oxygen fluxes within a mosaic of Fe limitation in the California Current Upwelling System
Abstract. We compare primary productivity estimates based on different photosynthetic ‘currencies’ (electrons, O2 and carbon) collected from the dynamic coastal upwelling waters of the California Current. Fast Repetition Rate Fluorometry and O2/N2 measurements were used to collect high-resolution underway estimates of photosynthetic electron transport rates and net community productivity, respectively, alongside on-station 14C uptake experiments to measure gross carbon fixation rates. Our survey captured two upwelling filaments at Cape Blanco and Cape Mendocino with distinct biogeochemical signatures and iron availabilities, enabling us to examine photosynthetic processes along a natural iron gradient. Significant differences in photo-physiology, cell sizes, Si:NO3- draw-down ratios, and molecular markers of Fe-stress indicated that phytoplankton assemblages near Cape Mendocino were Fe-stressed, while those near Cape Blanco were Fe-replete. Upwelling of O2-poor deep water to the surface complicated O2-based net community productivity estimates, but we were able to correct for these vertical mixing effects using continuous [N2O] surface measurements and depth-profiles of ∂[O2]∂[N2O]. Vertical mixing corrections were strongly correlated to sea surface temperature, which serves as an N2O-independent proxy for upwelling. Following vertical mixing corrections, all three productivity estimates reflected trends in Fe-stress physiology, indicating greater productivity near Cape Blanco compared to Cape Mendocino. For all assemblages, carbon fixation varied as a hyperbolic function of electron transport rates, but the derived parameters of this relationship were highly variable and significantly correlated with physiological indicators of Fe-stress (σPSII, FV/FM, Si:NO3- and diatom-specific PSI gene expression), suggesting that iron availability influenced the coupling between photosynthetic electron transport and subsequent carbon fixation. Net community productivity showed strong coherence with daily-integrated photosynthetic electron transport rates across the entire cruise track, with no apparent relationship with Fe-stress. This result suggests that fluorescence-based estimates of gross photochemistry are still a good indicator for bulk primary productivity, even if Fe-limitation influences the stoichiometric relationship between productivity currencies.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(4712 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4712 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-3812', Anonymous Referee #1, 04 Feb 2025
-
AC1: 'Reply on RC1', Yayla Sezginer, 26 Mar 2025
We thank reviewer 1 for their positive feedback and edits, which strengthen this study. Reviewer 1's suggestions are in bold and our responses are listed below. All edits are referenced by line number to the corresponding file of manuscript track changes.
L90: in the list of respiration pathways for NCP, shouldn’t bacterial respiration as well as respiration by larger animals be added to this list? See also Figure 7.
We have amended the statement to ‘community wide’ respiration to cover respiration by respiration by bacteria and macrofauna. L110 in track changes file.
Figure 1: panels a and b are reversed in the caption.
Corrected, L143-144.
L151: was PAR measured or assumed? If it was measured, how was it measured?
Here, PAR is delivered by instrument LED lamps that provide continuous actinic light. PAR provided by the instrument was not directly measured during ETRPSII measurements, but the instrument LEDs were calibrated against a WALZ ULM-500 PAR meter prior to deployment. Note on calibration added L180
Eq 4: what is meant by “in situ PAR”? PAR averaged for the MLD? I’m struggling to make sense of the equation.
Yes, PARin-situ refers to the mean PAR in the mixed layer. This equation closely follows that Eq. 3 in Domingues and Barbosa (2023), although some of the parameter names differ between our manuscript and their paper to maintain consistency within our manuscript. For example, here we denote the mixed layer depth as MLD, while Dominges and Barbosa use ‘Zm’. Some text added for clarification (L214).
Eq 7: the superscript “B” – is it a designation (bio?) or an exponent? Can you make it a bit clearer by spelling it out?
Yes, B for Bio. Clarified text reads as follows: “Biological concentrations, indicated by the superscript, ‘B’, are derived by isolating and removing physical solubility effects from measured gas concentrations.” (L316).
L262: inconsistency regarding letter k in the fraction: it’s capitals here, but small-caps above (L253).
Corrected (L329)
L266: since it’s the liquid fraction you’re interested in, should it be “filtered through” rather than “onto”?
Corrected (L361).
Figure 2: could you indicate here what areas were used for the boxplots in Figure 3?
New panel added to Figure 3 to indicate data allocation between subregions.
Note that although some ‘Cape Blanco’ data between 42.5-41.5N are at a relatively western longitude they are designated part of the coastal upwelling plume based on their low SST. Subregions were identified according to latitude and temperature criteria. Samples were considered part of an upwelling group if they had SST < 12 oC. If upwelling samples were north of 41N they were considered part of the Cape Blanco plume. If south of 41N, they were designated Cape Mendocino samples. All other samples were grouped in the offshore category. Criteria were determined by studying Figure 1a.
L365-368: These statements could distinguish a bit more between the capes – not all statements are true for both capes (Figure 4).
In reviewing Figure 4, there is evidence of diel cycles in Fv/Fm, , and NPQNSV. However, it is true that the magnitude of apparent diel cycles varies between Cape Blanco and Cape Mendocino. To acknowledge that, we have added the statement, ‘the magnitude of diel variability in FV/FM, , and NPQNSV signals displayed significant variability between subregions, as discussed below’ (L489). In-depth analysis of differences in photo-physiology between the two capes is discussed in the following paragraphs and Table 1.
Additionally, there is some convolution of the diel signals since the ship typically stayed close to the shelf during the day and transited offshore overnight. We have already acknowledged this in 507-514; ‘We note, however, that there is potential for some convolution of temporal and spatial variability, as the ship spent more time offshore in the night, and on-shore during the daytime. It is thus possible, that some of the diel cycling partially reflects different photo-physiological signals between coastal and offshore waters.’
L378/79: this statement is not true for alpha at Cape Blanco – this whole section could be either a bit more refined, or throw in a few more “overall”s to indicate that what is said is not always true but a general trend.
Threw in some generalizing terms: ‘Generally’ (L484), ‘Overall’ (L501)
L380: is beta shown anywhere?
Beta is included in Table 1 now.
L387: “Elevated values…” – this is only true for Cape Blanco. You said previously that both capes are upwelling, but to different degrees; maybe start early with a clear distinction between the two capes, but “upwelling” is not one of them, that’s true for both (see also L389).
Updated the text to reflect the general trends and relationships reported in Table 1 between upwelling conditions and photo-physiology, while still setting up the reader to understand how/why photo-physiology can still differs between two upwelling areas.
L516-523:
In general, FV/FM, Pmax, Ek and displayed positive relationships with upwelling indicators, i.e. salinity, nitrate, and decreased sea surface temperature (Table 1), suggesting that vertical transport of nutrient-rich water to the surface supported high photochemical yields. In contrast, signs of upwelling were associated with decreased and NPQNSV. However, despite general trends between photo-physiological parameters and upwelling, there were significant differences in photo-physiological properties between the Cape Blanco and Cape Mendocino upwelling plumes. At Cape Blanco…
L416/17: Is this the “in situ PAR” that confused me above? If so, make sure to call it that in the Figure.
Yes, added to figure 5a legend and caption.
L425-429: refer to Figure 5 here?
Done (L579).
L440-444: maybe add a word here about the uncertainties (due to nitrification in the surface layer) that were mentioned in the methods? What order of magnitude are they?
Added an estimate of error due to euphotic zone nitrification and added some details to the methods describing how this error was determined.
L331-341:
“We note that several recent studies have observed nitrification within the euphotic zone, challenging the assumption that N2O production is limited to subsurface waters (Grundle, Juniper and Giesbrecht, 2013; Smith et al., 2014), and potentially leading to overestimates in our vertical mixing-corrected NCP estimates. Previous observations in the CCS reported a range of depth-integrated mixed layer nitrification rates between 0.3 – 2 mmol NH4+ m-2 d-1, resulting in consumption of 0.6 – 4 mmol O2 m-2 d-1 (Stephens et al., 2020). Following the approach of Izett et al. (2018) we used a range of realistic N2O:O2 stoichiometries to estimate potential upper and lower bounds of mixed layer N2O production. We determined mixed layer N2O production likely ranged between 0.09 – 0.23 mol N2O m-2 d-1, which would yield offsets in our final NCP estimates between 1.2 and 3.3 mmol O2 m-2 d-1. Total uncertainty due to sources of error in other derived parameters was determined by following Izett (2021).
L594 – 596:
“Uncertainty in vertical-mixing corrected NCP due potential mixed layer nitrification (see sect. 2.6) represented between 1.5 – 4.2% of our mean corrected NCP value.”
L451: “x+/-y” wants to be filled with numbers
Yikes! Amended with the values.
L608-609: “On average, the bottom depth of the euphotic zone was 14 12 meters deeper than the bottom depth of the mixed layer”
Figure 7 and its caption: Do you mean “spatial and time scales”? Only the time scales are mentioned in the Figure. Also, I can’t see pathway 7 anywhere in the Figure
Removed ‘spatial’ and added a 7 symbol in the NCP panel.
L633: Did you mean to refer to Figure 6 here (rather than Figure 8)?
Removed the reference to figure 8, which does not belong here. L791
L653: “… with low Cmax” – this should be “…with low Pmax-C”, no?
Corrected.
Figure 8 and its caption: maybe make it clearer in the caption that blue and red also refer to different Fe-limitation regimes, or make a second panel where the dots are coloured by some measure of Fe limitation (Fv/Fm, sigma,… something like that)? This is a useful distinction, no? Also, Vmax (mentioned in the caption) is not shown, nor do I know what it would be based on what has been discussed?
Vmax in caption corrected to Pmax-C. Added some text to the Figure 8 caption.
L664: mention of Vmax again – should this be Ksat? It’s getting a bit confusing in this paragraph… if there is both a Vmax and a Ksat, then maybe a table of all the variables and their meaning would be useful?
Apologies for the confusion between Vmax, Pmax-C and Cmax. Pmax-C is the maximum Chl-normalized carbon fixation rate. The parameter pmax-C is comparable with Vmax in traditional Micahelis-Menten terminology. However, the maximum carbon fixation rate, whether expressed as a function of light, as in a PI curve, or as a function of ETR is nearly the same (no difference in the maximum measured carbon fixation rate, but the derived value changes very slightly between models). When writing this manuscript, there were many iterations where we tested different parameter names for Pmax-C, and some of the old names, Cmax, Vmax, etc, were not caught during editing. Thank you for your careful attention to this detail! Section 4.3.1 has been carefully reviewed now to catch all mentions of Cmax/Vmax/or Pmax-C and make sure they all read Pmax-C.
L696/70: I would have thought that Cyt b559a would correlate with NPQ, and therefore expected a negative correlation with Pmax-C. Does this deserve a bit more discussion?
I would agree that cyt b559a would be expected to positively correlate with NPQ, however I would disagree that Pmax should be negatively correlated with NPQ. Samples with high and low Pmax-C both expressed high levels of NPQ. One of the main findings of this section is that in some cases, carbon fixation cannot be predicted from NPQ. The text has been modified as follows:
While the precise functional roles of Cyt b559a are still not certain, previous studies have demonstrated its potential role in photoprotective cyclic electron transport around PSII and PSII assembly (Chiu and Chu, 2022). L952-954.
L678: Vmax again… and also, a question that comes up reading this paragraph and the preceding ones: How do you decide whether to discuss Vmax (Ksat) rather than Pmax-C? Is there a scheme to this? They seem somewhat interchangeable to me at this point. When you don’t mention one or the other, does it mean the correlation would have been insignificant? Again, maybe a word or two on this would be useful.
Text around here has changed quite a bit so that text is gone now (see following comment). However, the point about interchangeable parameters is still a good one.
The text has been cleaned up so only Pmax-C is mentioned throughout. Kssat and Pmax-C are not interchangeable. Rather, the ksat and Pmax-C terms discussed here are comparable with the Ksat and Vmax terms in a Michealis-Menten equation.
Discussion is mostly limited to Pmax-C for simplicity and because this parameter is more representative of photosynthetic enzymes concentrations and kinetics, which provides mechanistic insights into photosynthetic functioning and are useful to understand the observed variability in e:C/. It’s worth noting that Kssat and Pmax-C are positively correlated ( = 0.70, p << 0.01) L954-956.
L678-80: This sentence is either missing a final section, or the “and” after the comma needs to go; also, please specify whether this increased coupling of Chl to RCII is specific to iron limitation or generally true.
Lots of new details added to the discussion of Chl:RCII and its influence on our comparison of carbon fixation and electron transport rates:
“In addition to non-linear electron transport, is also directly affected by the number of Chl energetically coupled to RCII (. Directly measuring requires specialized O2 flash yield instrumentation (Suggett et al., 2009), which was unavailable for this study. However, in-situ Chl concentrations, normalized to FRRF-derived proxies for [RCII] (∝ Fo/) following the approach of Oxborough et al. (2012), can be used to examine variability in between samples. With a known instrument calibration factor, Ka, either provided by instrument manufacturers or determined independently by O2 flash yield measurements, this approximation could be used to estimate the absolute value of .
It is well established that Chl:RCII () ratios increase under low light, to maximize light absorption (Greenbaum and Mauzerall, 1991). In our measurements, the proxy for varied significantly between sample depths, with higher at the bottom of the euphotic zone compared to surface depths, confirmed by a t-test comparison of population means (p << 0.01). Iron limitation is also expected to increase Chl:RCII. Although iron limitation lowers total cellular Chl content, Chl is more likely to be energetically coupled to RCII rather than PSI reaction centers (Greene et al., 1992). Accordingly, displayed a negative correlation with Fea1 expression in surface samples ( = -0.72, p <0.05, n = 9), which we used as a proxy for iron limitation. We thus conclude that Fe-stress likely contributed to variability in in addition to influencing non-linear electron transport. The hyperbolic relationship between carbon fixation and electron transport was unaffected by , which was assumed to be constant for individual samples throughout the course of photosynthesis-irradiance experiments (Appendix 4). However, this assumption may be violated under high light, due to photoinactivation of RCII (Campbell and Serôdio, 2020). A robust understanding of variability requires direct [RCII] measurements collected in parallel with ETRPSII and carbon fixation measurements.”
L681: “increases in”? or “increased”? This whole paragraph could be a bit clearer - what does the literature say, what was measured here, what needs to be assumed, how does it all come together?
This text has been replaced with the paragraphs above.
L704: Can you insert the actual ranges that have been observed by others?
Kranz et al. (2020) observed GPP rates of 0 – 4000 mmol C m-2 d-1, 0 – 4000 mmol O2 m-2 d-1 from O2/Ar- and FRRF- based approaches. This range is coincident with that observed in this study (L1047).
Paragraph starting L698: The observed relationship between FRRf-derived GPP and NCP is a significant and surprising result (maybe you disagree, but then please lay out more clearly why it’s not surprising). Maybe a bit more discussion on why this relationship with NCP works here would not be amiss. Where/when do we expect this to work elsewhere (and where/when may it fall over)? What needs to be investigated next? Does it have anything to do with iron at all?
Agreed! This is a significant and perhaps surprising result, due to the large differences in integration time scales of GPP and NCP, and the potential for different metabolic and ecological processes to decouple these rates (L918).
We initially hypothesized that ETR:NCP would vary with ETR:C-fixation, and that factors, like Fe availability, that drive decoupling between ETR and GPP would similarly decouple ETR and NCP. However, we did not observe significant differences in ETR:NCP between Cape Blanco and Cape Mendocino as would be expected if Fe was playing a key role in ETR:NCP variability. Rather, the relationship between ETR:NCP was fairly consistent across the study region. We offer an alternative explanation for the surprising consistency in the ETR:NCP relationship below (L1093-1113):
Despite the inherent dependency of net oxygen production on photosynthetic electron transport rates, the strong correlation between NCP and ETRPSII and the consistency of ETRPSII:NCPacross our entire study regions is surprising given the large number of methodological and physiological factors that can significantly uncouple these rates (Fig 7). However, NCP and GP estimates both have similar dependencies on mixed layer Chl concentration. To obtain FRRF-derived GP estimates in comparable units of mmol O2 m-2 d-1, we multiplied in-situ ETRPSII by mixed layer Chl concentrations (Eq 5). Although mixed layer Chl concentrations are not explicitly included in NCP calculations (Sect 2.6), biomass is expected to be a primary driver of bulk productivity. Indeed, when we compared Chl-normalized ETRPSII and NCP estimates, we found a much weaker relationship between with correlation coefficients decreasing to 0.22 and 0.35 for 24h binned and instantaneous measurements, respectively. We therefore conclude that it remains challenging to derive gross and net carbon fluxes from FRRF measurements alone, but paired ETRPSII and Chl measurements can provide useful constraints for NCP estimates.
Figure 9 and caption: Does the correlation for panel b (rho = 0,92,L715) include the negative data point? How should one think about the negative data point, can it be explained, could it have to do with upwelling/downwelling? And what about the intercept for the line of best fit (-0.55), how to think about that? That it’s very close to zero and that’s good, or does it have some other meaning that’s worth spelling out? Also, the final sentence of the figure caption is a repeat, except for the r2. Regarding panel A of the figure: could it be improved by plotting the points as a heat map, to show where most of the data points fall?
The correlation does include the negative data point. This point corresponds to measurements on June 2. On this day NCP reached its minima. This point corresponded with high SST indicative of offshore waters outside the upwelling plume (see Fig 5b).
The y-intercept being close to zero is a positive indication that our 24h binned NCP and ETR comparison are reasonable, however this y-intercept should not be taken as an absolute. As we can see from the highly negative point in Fig 9b, very negative NCP values are possible in net heterotrophic waters, while ETR values can never be net negative.
Repeat sentence removed from Fig 9 caption
A heat map of the data density indicates that most of the data is concentrated around [ETR = 0, ncp = 100]. This makes sense since ~half the ETR data should = 0 overnight and the mean ncp was about 80 mmol O2 m-2 s-1. Measurements collected during the day (ETR > 0) appear evenly spread. Since the point that ETR must equal 0 overnight is made a few times in the text, we have decided not to change the manuscript in the figure to the heat map, but hope it is of interest for Reviewer 1.
L739: Regarding “fine-scale variability”: what scales are being resolved when we go to 24h binning?
This is an excellent point! A primary motivation for using FRRF and O2/N2 is the ability to capture fine-scale variability. That fine-scale variability is of course smeared if a 24h bin is applied. The spatial ‘smear’ will depend on the boat’s movement throughout the day. In our case, the boat stayed on-station during the day for full-scale morning and afternoon sampling programs and transited to the following station overnight.
Here each 24h binned ETR measurement roughly represents the mean photochemical activity at each station over 24 hours, since ETR is zero overnight when the ship was transiting. The interpretation is more complicated for NCP which captures variability in O2/N2’ overnight due to variable respiration rates as the ship transited through different water masses.
Binning obviously negates the critical advantages of high frequency measurement systems like the FRRF and PIGI. However, we applied this approach to purposefully smear the temporal resolution of ETR to approximate the time-scales represented by O2/N2’-based NCP measurements and provide a more just comparison. This approach is far from perfect, but we found it useful to illustrate how trying to account for differences in time-scales can improve the correlation between productivity metrics.
Citation: https://doi.org/10.5194/egusphere-2024-3812-AC1
-
AC1: 'Reply on RC1', Yayla Sezginer, 26 Mar 2025
-
RC2: 'Comment on egusphere-2024-3812', Kazuhiro Yoshida, 28 Feb 2025
Review Report
Sezginer et al. estimated primary productivity in two different upwelling areas with multiple approaches for photosynthesis measurements. It has been controversial but also dreamt estimation of primary production by measuring real-time and/or autonomous chl fluorescence measurements; however, few studies have addressed the discrepancies between fluorescence and carbon fixation/oxygen evolution particularly in natural waters. The electron-carbon conversions were successful in restricted and homogenous waters (e.g., Cheah et al. 2011). Sezgniger et al. comprehensively compared FRRf, O2, and carbon on a large scale within different Fe-mosaiced upwelling systems. They did not just measure primary production; they also mentioned underlying transcriptional changes along with different photosynthetic performance and Fe availability. Overall, they clarify the hope and limitations of the dream bridge between electron-oxygen-carbon. Although the manuscript reads well, I suggest further analysis or clarifications of your interpretation of the data. I thus recommend Moderate revision before publication.
[Suggestions]
A few mentions of phytoplankton taxonomy
- During your sampling, diatoms, particularly Cheatoceros, dominated the phytoplankton communities in both upwelling areas. Undoubtedly, diatoms were most abundant in your samples; however, you might also observe other diatoms, dinoflagellates and/or small taxa. You conducted light microscopy to look at the phytoplankton taxonomy of your samples. If your microscopy data was sufficiently reliable, subsection 4.2 can be improved as Taguchi (1976), Finkel (2000), and Suggett et al. (2009) reported that different taxa showed unique photosynthetic performances even both carbon and fluorescence.
More discussion on Carbon/nitrogen ratio
- I agree Fe availability modifies the photosynthesis machinery such as PSI: PSII and Cytochrome: PSII, which decouples the electron: carbon conversion. However, photosynthetic energy can be allocated to both carbon and nitrogen et cetera, which could also decouple the conversion factor. You only discuss C: N ratios compared with the Redfield ratio. Discussion on the relative allocation to nitrogen from C: N ratios can improve your discussion on the decoupling electron: carbon.
Derivation of qP or qL
Your discussion on low Fe effects on FRRf and carbon fixation was on non-linear electron transport due to an imbalance between photosystems. I understand you well discuss the reduction of PSI, inferred from TPMs (or relative gene expression) of the PSI-related genes. In addition, I would suggest you calculate 1-qP or 1-qL as the state of open/closed PSII from your FRRf data.
How much RNA data did you throw away?
I agree your transcriptomics indeed deepened your discussion. I have two technical questions:
(1) Have you compared with other de novo assemblers than rnaSpades or CD-HIT-EST? Any assembling performance data? Did you pre-process raw RNA sequences?
- (2) You extracted diatom reads/contigs for further downstream analysis. However, transcriptomic data might be unbiased and might have various reads/contigs from various organisms. What proportion did diatom data contribute to the total sequences? How much RNA data did you throw when only focusing on diatoms?
[Minor issues]
L110: Have you validated your nitrate sensor with your autoanalyzer outputs?
L139: How did you dissimilate/separate QA and QB reoxidation from your FRRf data? You can separate QA and QB if using an algal isolate, but was it appropriately separated in natural communities even with various Fe availabilities?
L203: I understand NPQ fitting is quite tricky. Serôio and Lavaud (2011) thus discussed model performance on NPQ-E fittings. How were your single-component exponential fittings? Any stats outputs?
L206: Why did you choose 46% light depth? You surely collected the same light depths at each station? Or simply light intensities at ~50 m almost similar at each station?
L250: Any other source(s) of N2O other than photoinhibition?
L321 and elsewhere: How did you measure salinity? If from CTD, psu is no longer used. Just unitless.
L354: In general, the Redfield ratio does not apply to surface waters I reckon because Redfield ratio 16 was measured in deep waters. I would rather recommend C: N allocation ratios in algal cells as discussed above.
Figure 7: As you discuss, Cape Bianco had lower temperatures due to stronger upwelling. It is quite interesting but did this temperature difference affect (1) the enzymatic activity of xanthophyll pigment synthesis and consequently NPQ dynamics/xanthophyll cycle (XC)? and (2) transcriptomic regulation such as low-temperature adaptation, which overlayed Fe-regulated gene expression or TPM counts?
Regarding (1), NPQ correlated with SST in Table 1, which might be due to temperature-dependence XC or NPQ? Also, differences in the community structure might influence NPQ or other FRRf parameters?
Table 1: In the 3rd row, Pmax should be ETRmax??
Figure 5(c): GP at Cape Bianco is highly variable with numerous spiky up-downs compared with the peaceful Cape Mendocino. Any possible explanations for the difference? Or simply, large errors for estimation of GP in Cape Bianco?
L451: I wanna know x and y! Please add the exact numbers.
Figure 6: Quite interesting! I’m curious Stations 2 and 9 showed almost the same conversion factors between surface and subsurface. Any possible explanations? Because of intense vertical mixing at both stations?
Subsections 4.2 and 4.3 are a bit redundant for me even as an aquatic photosynthesis researcher although your discussion is quite interesting! It would be great if you could shorten these paragraphs.
Subsection 4.3.2: There is no mention of Fe availability. No need here? It would be more interesting if you could include just a few discussions on Fe availability as your manuscript title suggests.
L685: OK, I would partially agree with your rough estimation using 400-700 chls/RCII. Is it also applicable to Fe-limited waters such as Cape Mednonico? If you look through other chls/RCII ratios in Fe-limited waters (e.g., Strzepek et al. 2019 with appendices and references therein).
L731: Energy imbalance due to photosystem impairment can be indeed inferred from NPQNSV as you discuss. I would also suggest looking at qP or qL as described above.
Review Report
Sezginer et al. estimated primary productivity in two different upwelling areas using multiple approaches for photosynthesis measurements. The estimation of primary production by measuring real-time and/or autonomous chl fluorescence has been controversial but also considered a dream; however, few studies have addressed the discrepancies between fluorescence and carbon fixation/oxygen evolution, particularly in natural waters. Electron-carbon conversions have been successful in restricted and homogeneous waters (e.g., Cheah et al. 2011). Sezginer et al. comprehensively compared FRRf, O₂, and carbon on a large scale within different Fe-mosaiced upwelling systems. They did not only measure primary production; they also examined underlying transcriptional changes along with different photosynthetic performances and Fe availability. Overall, they clarify both hopes and limitations of the dream bridging between electron-oxygen-carbon. Although the manuscript reads well, I suggest further analysis or clarifications of your data interpretation. I, therefore, recommend moderate revision before publication.
[Suggestions]
- A few mentions of phytoplankton taxonomy
During your sampling, diatoms, particularly Chaetoceros, dominated the phytoplankton communities in both upwelling areas. Undoubtedly, diatoms were the most abundant taxa in your samples; however, you might have also observed other diatoms, dinoflagellates, and/or small taxa. You conducted light microscopy to examine the phytoplankton taxonomy of your samples. If your microscopy data were sufficiently reliable, subsection 4.2 could be improved, as Taguchi (1976), Finkel (2000), and Suggett et al. (2009) reported that different taxa exhibit unique photosynthetic performances in both carbon fixation and fluorescence. - More discussion on Carbon/Nitrogen ratio
I agree that Fe availability modifies the photosynthesis machinery, such as PSI:PSII and Cytochrome:PSII, which decouples electron:carbon conversion. However, photosynthetic energy can be allocated to both carbon and nitrogen, among other processes, which could also decouple the conversion factor. You only discuss C:N ratios compared with the Redfield ratio. A discussion on the relative allocation to nitrogen based on C:N ratios could improve your analysis of the decoupling between electron and carbon conversion. - Derivation of qP or qL
Your discussion on the effects of low Fe on FRRf and carbon fixation focuses on non-linear electron transport due to an imbalance between photosystems. While you provide a well-developed discussion on PSI reduction, inferred from TPMs (or relative gene expression) of PSI-related genes, I suggest calculating 1-qP or 1-qL to assess the state of open/closed PSII from your FRRf data.
- How much RNA data did you discard?
I agree that your transcriptomic analysis significantly deepened the discussion. However, I have two technical questions:
(1) Have you compared other de novo assemblers besides rnaSpades or CD-HIT-EST? Do you have any assembly performance data? Did you pre-process raw RNA sequences?
(2) You extracted diatom reads/contigs for further downstream analysis. However, transcriptomic data are typically unbiased and may include reads/contigs from various organisms. What proportion of the total sequences was attributed to diatoms? How much RNA data did you discard when focusing only on diatoms?
[Minor issues]
L110: Have you validated your nitrate sensor against your autoanalyzer outputs?
L139: How did you distinguish/separate QA and QB reoxidation from your FRRf data? While separation is possible with an algal isolate, was it appropriately achieved in natural communities with varying Fe availabilities?
L203: I understand that NPQ fitting is quite tricky. Serôdio and Lavaud (2011) discussed model performance on NPQ-E fittings. How well did your single-component exponential fittings perform? Any statistical outputs?
L206: Why did you choose 46% light depth? Did you collect samples at the same light depth for each station, or were light intensities at ~50 m similar across stations?
L250: Were there any additional sources of N₂O other than photoinhibition?
L321 and elsewhere: How did you measure salinity? If you used a CTD, note that PSU is no longer used and should be reported as a unitless value.
L354: The Redfield ratio generally does not apply to surface waters, as it was derived from deep-water measurements. I suggest discussing C:N allocation ratios in algal cells instead.
Figure 7: As you noted, Cape Bianco had lower temperatures due to stronger upwelling. Did this temperature difference influence (1) the enzymatic activity related to xanthophyll pigment synthesis and, consequently, NPQ dynamics/xanthophyll cycling (XC), or (2) transcriptomic regulation, such as low-temperature adaptation, which could have overlaid Fe-regulated gene expression or TPM counts? Regarding (1), NPQ correlates with SST in Table 1. Could this be due to temperature-dependent XC or NPQ? Additionally, could differences in community structure have influenced NPQ or other FRRf parameters?
Table 1: In the third row, should Pmax be labeled ETRmax instead?
Figure 5(c): GP at Cape Bianco is highly variable, with numerous spiky ups-downs compared to the stable and peaceful Cape Mendocino. What might explain this difference? Or is it simply due to large estimation errors for GP at Cape Bianco?
L451: I wanna know x and y! Please provide the exact values for x and y.
Figure 6: Very interesting! Stations 2 and 9 showed nearly identical conversion factors between surface and subsurface. Any possible explanations? Could intense vertical mixing at both stations be responsible?
Subsections 4.2 and 4.3: These sections feel somewhat redundant, even for an aquatic photosynthesis researcher like myself. While your discussion is quite interesting, consider shortening these paragraphs for clarity.
Subsection 4.3.2: There is no mention of Fe availability here. Is this intentional? Including a brief discussion on Fe availability would make this section more aligned with your manuscript’s title.
L685: I partially agree with your rough estimation using 400-700 chls/RCII. However, is this also applicable to Fe-limited waters such as Cape Mendocino? I recommend reviewing other chl/RCII ratios in Fe-limited waters (e.g., Strzepek et al. 2019, including appendices and references therein).
L731: As you discuss, energy imbalance due to photosystem impairment can be inferred from NPQNSV. Additionally, I suggest considering qP or qL, as described above.
Citation: https://doi.org/10.5194/egusphere-2024-3812-RC2 -
AC2: 'Reply on RC2', Yayla Sezginer, 26 Mar 2025
Publisher’s note: this comment was edited on 7 April 2025. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.
We similarly thank Reviewer 2 for their positive feedback, expert insights, and suggestions for clarification and deeper analysis. We have aimed to incorporate as many of their suggestions as possible. Unfortunately, not all the datasets they have requested for deeper analysis are available. Where further data analysis is not possible, we strive to incorporate the additional considerations Reviewer 2 suggests in our discussion. Again, Reviewer 2’s suggestions are listed below in bold with our beneath. Line references in our replies correspond to the lines in the updated manuscript with track changes.
A few mentions of phytoplankton taxonomy
- During your sampling, diatoms, particularly Cheatoceros, dominated the phytoplankton communities in both upwelling areas. Undoubtedly, diatoms were most abundant in your samples; however, you might also observe other diatoms, dinoflagellates and/or small taxa. You conducted light microscopy to look at the phytoplankton taxonomy of your samples. If your microscopy data was sufficiently reliable, subsection 4.2 can be improved as Taguchi (1976), Finkel (2000), and Suggett et al. (2009) reported that different taxa showed unique photosynthetic performances even both carbon and fluorescence.
Our taxonomic analysis primarily relied on chemotaxonomic analysis of pigment data which does not differentiate between diatom taxa. Microscopic analysis was limited to surface samples and was useful to validate our chemotaxonomic analysis and identify general trends between stations, i.e. presence of Pseudo-Nitszchia at Cape Mendocino, but not Cape Blanco. However, this analysis is pretty biased towards larger more recognizable cells and was used more qualitatively. This data is likely not sufficient to apply robust statistical analyses to evaluate taxonomically driven variability in photo-physiology. This is unfortunate since, as you say, it is clear from the literature that different taxa express different optical phenotypes influencing Chlorophyll fluorescence signatures. In a field study, you could argue that variability in these optical phenotypes still ultimately derives from the environmental conditions that select for different phytoplankton taxa.
More discussion on Carbon/nitrogen ratio
- I agree Fe availability modifies the photosynthesis machinery such as PSI: PSII and Cytochrome: PSII, which decouples the electron: carbon conversion. However, photosynthetic energy can be allocated to both carbon and nitrogen et cetera, which could also decouple the conversion factor. You only discuss C: N ratios compared with the Redfield ratio. Discussion on the relative allocation to nitrogen from C: N ratios can improve your discussion on the decoupling electron: carbon.
Agreed, N-assimilation is a decoupling pathway that is not directly triggered by excess energy and could therefore help explain why NPQ and e:C are not always correlated. Further, the source of N is also likely to affect e:C as growth on more reduced forms (i.e. NH4) will divert less e from carbon fixation. However, N-assimilation is also negatively impacted by Fe-limitation as both nitrate and nitrite reductase use Fe as cofactors. Allen et al. (2008) demonstrated these enzymes were downregulated under Fe limitation. While previous studies of diatoms have found either no change in cellular N:P or decreases in N:P in response to Fe limitation (Price 2005; La Roche et al., 2003). As a result, it is not clear how C:N may have varied between our Fe rich and Fe stressed study sites, or how Fe limitation may have influenced e:C decoupling through nitrate assimilation. This is an important point to resolve for the community moving forward. Unfortunately, without the accompanying data, we feel this topic is somewhat out of the scope of our paper.
Derivation of qP or qL
Your discussion on low Fe effects on FRRf and carbon fixation was on non-linear electron transport due to an imbalance between photosystems. I understand you well discuss the reduction of PSI, inferred from TPMs (or relative gene expression) of the PSI-related genes. In addition, I would suggest you calculate 1-qP or 1-qL as the state of open/closed PSII from your FRRf data.
Yes, qP (fraction of open RCII) plays an important role in determining ETRPSII, and is actually included in our ETRPSII equation (Eq.1) where it is shown under its other notation, F’q/F’v (see Tortell, Suggett and Schuback 2023 for a list of synonymous FRRF nomenclature). While we did discuss potential Fe effects on F’q/F’v (L752-758), we did not take the next steps in actually analyzing our F’q/F’v data. Following Reviewer 2’s suggestion, we have now actually examined qP along the cruise transect (see below).
The figure above displays the parameter 1-qP. qP was measured at light levels ranging from 0-850 uE during photosynthesis-irradiance curves. To estimate the fraction of closed RCII, in-situ, we applied the same approach we used to determine underway NPQ, where 1-qP was plotted against PAR, and fit with an exponential curve (92% of model fits had R2 > 0.9). In-situ 1-qP was then estimated by inputting in-situ PAR into the resulting model equation. Our results demonstrate strong diel patterns due to the immediate dependence of this approach on PAR, but also highlight that the fraction of RCII closure during peak midday irradiance was typically lower for Cape Blanco samples (5/30 – 06/04) compared to Cape Mendocino (06/06 – 06/10), with the exception of data collected on 06/02, which is when we transited offshore (note, this is also where we recorded our lowest NCP measurement).
To minimize confusion and avoid adding a new parameter name in the text, we are opting not to use the qP terminology and remain consistent with F’q/F’v. We have added analysis of F’q/F’v to Table 1, and added a note to our discussion section (L770-772):
‘Indeed, measured during underway Photosynthesis-Irradiance curves and mapped onto in-situ irradiances, demonstrated that was higher around Cape Blanco compared to Cape Mendocino (Table 1)’
We also added some text to the methods (Section 2.4) to explain how we evaluated in-situ F’q/F’v. (L256-259)
How much RNA data did you throw away?
Please see the table below contributed by coauthor Emily Speciale detailing our sequencing stats.
Station
Rep
Total Reads
# Reads Mapped
% Reads Mapped
# Reads Taxonomically Annotated
% Reads Taxonomically Annotated
# Protist Group
% Protist Group
# Diatom
% Diatom
1
A
20919052
15566516
74.41
12163291
58.14
8529884
40.78
4462784
21.33
B
20605487
14828573
71.96
11629423
56.44
8495042
41.23
5035031
24.44
2
A
20370660
16757515
82.26
12441421
61.08
9060238
44.48
3337286
16.38
B
33411917
23851899
71.39
17576726
52.61
3878516
11.61
1577339
4.72
C
20451261
16318259
79.79
13878818
67.86
6014888
29.41
3218519
15.74
3
A
31854033
24637176
77.34
17205795
54.01
8931444
28.04
1585515
4.98
B
23102631
18257440
79.03
12937513
56.00
7632094
33.04
1385720
6.00
C
22184188
9543967
43.02
5778682
26.05
1516238
6.83
601494
2.71
4
A
31244147
22594019
72.31
14833206
47.48
8988587
28.77
3345948
10.71
B
17121856
13681984
79.91
9435876
55.11
6813257
39.79
2557063
14.93
C
21784053
16405679
75.31
12470858
57.25
9922004
45.55
5794848
26.60
5
A
24551450
19993489
81.44
16862370
68.68
15686918
63.89
3354493
13.66
B
20633218
15952493
77.31
13714182
66.47
12762334
61.85
3482216
16.88
C
13037287
5936298
45.53
4437254
34.04
632351
4.85
253280
1.94
6
A
18605896
14369514
77.23
11475963
61.68
8299338
44.61
2787489
14.98
B
20757774
7441701
35.85
3668990
17.68
1359173
6.55
441358
2.13
C
22992753
18563929
80.74
14898492
64.80
10666864
46.39
4200265
18.27
7
A
13046360
10221886
78.35
6465869
49.56
4729515
36.25
819835
6.28
B
22985192
18162597
79.02
13223369
57.53
8650540
37.64
1678663
7.30
C
17752669
8665680
48.81
4882063
27.50
2656758
14.97
840107
4.73
8
A
28417725
21660636
76.22
12402752
43.64
8187261
28.81
1294051
4.55
B
18902138
15344332
81.18
9352687
49.48
5208950
27.56
901508
4.77
C
18139851
12084503
66.62
9296668
51.25
6835677
37.68
1183001
6.52
9
A
22252339
17619164
79.18
14355113
64.51
13217881
59.40
1699119
7.64
B
22274606
17295073
77.64
14039185
63.03
12776475
57.36
1600294
7.18
C
18592697
15045958
80.92
11457739
61.62
10642495
57.24
1281205
6.89
10
A
19985953
17084390
85.48
11958397
59.83
8132358
40.69
1241650
6.21
B
14932416
12288932
82.30
8724712
58.43
5509449
36.90
791170
5.30
C
19818129
11020988
55.61
6792417
34.27
2580105
13.02
628061
3.17
SUM
620727738.00
451194590.00
72.69
328359831.00
52.90
218316629.60
35.17
61379310.90
9.89
AVG
21404404.76
15558434.14
72.28
11322752.79
52.62
7528159.64
35.35
2116527.96
9.90
STD. DEV.
4905885.70
4678756.37
13.18
3827375.65
13.20
3751501.90
16.77
1490867.62
6.91
I agree your transcriptomics indeed deepened your discussion. I have two technical questions:
- Have you compared with other de novo assemblers than rnaSpades or CD-HIT-EST? Any assembling performance data? Did you pre-process raw RNA sequences?
Coauthors Adrian Marchetti and Emily Speciale are part of an OCB Meta-Eukmoics working group that is doing an intercomparison of metatranscriptomic methods for microbial communities. Based on the work we've done thus far, there are no clear trends between the performance of different de novo assemblers (whether that be rnaSpades, Megahit, Trinity, etc.). Thus, we chose to use rnaSPAdes and CD-HIT-EST due to our familiarity with the softwares and their efficiency. We did pre-process raw reads using TrimGalore and FastQC.
(2) You extracted diatom reads/contigs for further downstream analysis. However, transcriptomic data might be unbiased and might have various reads/contigs from various organisms. What proportion did diatom data contribute to the total sequences? How much RNA data did you throw when only focusing on diatoms?
See the table above for information regarding the proportion of transcriptomic data.
We chose to focus on diatoms because we believe they dominant functional group among the photoautotrophs in this ecosystem and there is rich literature offering sufficient references to interpret diatom transcriptomic data. There are definitely substantial non-diatom organisms in our metatranscriptomics data that map to non-autotrophic organisms. When trying to look at broad expression trends, trophic mode can be hard to filter for, so we filtered for Bacillariophyta (aka diatom) mapped reads to represent the dominant phytoplankton response to various environmental conditions.
L110: Have you validated your nitrate sensor with your autoanalyzer outputs?
The sensor was recently recalibrated by Seabird prior to the cruise. Immediately prior to the cruise and during the cruise itself we validated the sensor against standards of sodium nitrate dissolved in MQ. We have added a note of this to the methods section (L35-36).
L139: How did you dissimilate/separate QA and QB reoxidation from your FRRf data? You can separate QA and QB if using an algal isolate, but was it appropriately separated in natural communities even with various Fe availabilities?
The FRRF method we applied is a single-turnover method which relies on the assumption that our full sequence of sub-saturating excitation pulses and following relaxation sequence leads to a single reduction and oxidation event of the collective (mixed assemblage) Qa pool. This assumption is based on the timing of the excitation sequence (127 sub-saturating pulses delivered over 250 us) which outpaces the reoxidation of Qb= -> PQH2 -> Cyt b6f which occurs over 4-5 ms and represents one of the rate limiting steps of photosynthetic light reactions. Multiple turnover methods with longer excitation time-scales (200 – 10,000 ms; e.g. PAM fluorometers) do record fluorescence signals associated with Qb reoxidation (see figure below).
Image credit: Chelsea Instruments. The multiple turnover shows a temporary pause in the rise of fluorescence when Qa is saturated. As Qa is reoxidized, fluorescence continues to increase until the entire electron transport chain is saturated. This is part of the reason Fv/Fm measured by PAM fluorometry is higher than FRRF.
L203: I understand NPQ fitting is quite tricky. Serôio and Lavaud (2011) thus discussed model performance on NPQ-E fittings. How were your single-component exponential fittings? Any stats outputs?
Stats added (L253): ‘Out of 91 curve fits, 95 % had R2 > 0.90 and 87% had R2 > 0.95.
A sample NPQ-PAR curve is included below.
L206: Why did you choose 46% light depth? You surely collected the same light depths at each station? Or simply light intensities at ~50 m almost similar at each station?
A primary objective of this cruise campaign was to collect measurements during a large scale Fe enrichment (+Fe) and depletion (+DFB) incubations. Light depths were sampled to mimic the on-deck incubators used for the Fe experiments. The Fe experiment is not part of this study, but influenced sampling strategies throughout the cruise for data comparison and compatibility. The bottom of the euphotic zone ranged between samples between 40-46m. Light depths were determined using the CTD-mounted PAR meter during our daily productivity casts enabling consistent sampling at 46% and 1% light levels.
L250: Any other source(s) of N2O other than photoinhibition?
Good Q, added some text to the methods and results section regarding uncertainty in NCP due to potential mixed layer nitrification, which was small.
L331-341: We note that several recent studies have observed nitrification within the euphotic zone, challenging the assumption that N2O production is limited to subsurface waters (Grundle, Juniper and Giesbrecht, 2013; Smith et al., 2014), and potentially leading to overestimates in our vertical mixing-corrected NCP estimates. Previous observations in the CCS reported a range of depth-integrated mixed layer nitrification rates between 0.3 – 2 mmol NH4+ m-2 d-1, resulting in consumption of 0.6 – 4 mmol O2 m-2 d-1 (Stephens et al., 2020). Following the approach of Izett et al. (2018) we used a range of realistic N2O:O2 stoichiometries to estimate potential upper and lower bounds of mixed layer N2O production. We determined mixed layer N2O production likely ranged between 0.09 – 0.23 mol N2O m-2 d-1, which would yield offsets in our final NCP estimates between 1.2 and 3.3 mmol O2 m-2 d-1. Total uncertainty due to sources of error in other derived parameters was determined by following Izett (2021).
L594-596: Uncertainty in vertical-mixing corrected NCP due potential mixed layer nitrification (see sect. 2.6) represented between 1.5 – 4.2% of our mean corrected NCP value.
L321 and elsewhere: How did you measure salinity? If from CTD, psu is no longer used. Just unitless.
Salinity was measured with a CTD and TSG. Psu units removed.
L354: In general, the Redfield ratio does not apply to surface waters I reckon because Redfield ratio 16 was measured in deep waters. I would rather recommend C: N allocation ratios in algal cells as discussed above.
Agreed that incorporating measurements of C:N in biomass (and Si) would have strengthened the manuscript. Unfortunately, that is not data we have on hand, so our discussion is limited to our measurements of dissolved N, P, and Si.
Figure 7: As you discuss, Cape Bianco had lower temperatures due to stronger upwelling. It is quite interesting but did this temperature difference affect (1) the enzymatic activity of xanthophyll pigment synthesis and consequently NPQ dynamics/xanthophyll cycle (XC)? and (2) transcriptomic regulation such as low-temperature adaptation, which overlayed Fe-regulated gene expression or TPM counts?
Regarding (1), NPQ correlated with SST in Table 1, which might be due to temperature-dependence XC or NPQ? Also, differences in the community structure might influence NPQ or other FRRf parameters?
Yes, NPQ, like Fv/Fm and other photophysiological parameters should vary with temperature and community structure. However, direct temperature effects on NPQ are expected to increase NPQ. Previous lab studies by Xu et al. (2012) (https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=phytoplankton+NPQ+cold+temperature&oq=phytoplankton+NPQ+cold+temper#d=gs_qabs&t=1743778443485&u=%23p%3DMa5utUAsm8gJ) have shown that NPQ increases under cold temperatures while While Yan et al. (2019) (https://onlinelibrary.wiley.com/doi/abs/10.1111/jpy.12846) demonstrated that Arctic phytoplankton enhanced NPQ photoprotective strategies under colder temperatures. Presumably, slower enzyme activity in the cold reduces capacity for protein repair and downstream electron transport, increasing susceptibility to photodamage or inhibition (Fanesi et al,.2016) (https://onlinelibrary.wiley.com/doi/abs/10.1111/fwb.12777). However, our data shows a positive correlation between NPQ and SST, opposite of the expected trend due to direct temperature effects. Here, SST acts as an indicator for nutrient-rich upwelling water. While the lower temperatures almost surely have an effect on enzymatic rates, it appears indirect sst effects, and therefore greater nutrient availability, are outweighing direct effects.
Table 1: In the 3rd row, Pmax should be ETRmax??
The Pmax refers to Pmax from photosynthesis-irradiance curves. Pmax is defined by eqn 3 in section 2.3 and the Pmax plotted in Fig 4b. Admittedly, Pmax/Vmax/Cmax terminology all got a bit muddled in section 4.3.1 and is likely the source of confusion here. All of the terminology throughout has been revised for consistency. Pmax is the maximum rate for ETR. Pmax-C is the maximum rate of carbon fixation.
Figure 5(c): GP at Cape Bianco is highly variable with numerous spiky up-downs compared with the peaceful Cape Mendocino. Any possible explanations for the difference? Or simply, large errors for estimation of GP in Cape Bianco?
Spiky data around May 31-June 1, indeed! Yet ETRPSII, which GP derives from is not spiky around those days. This indicates the variability is coming from the conversion factor from ETR to GP. To convert ETR to GP, we applied the following unit conversion:
Within the conversion factor above, the only variables that are not constant are the mixed layer depth and the mixed layer chlorophyll concentration. The chlorophyll data around Cape Blanco from May 31-June 1 was highly variable (see below) and that variability propagated into our GP estimates. Why was Chl data spikier around Cape Blanco compared to Cape Mendocino? Not sure exactly, but we were transiting across hydrographic fronts between offshore waters and the upwelling plume where a strong bloom was beginning to form.
L451: I wanna know x and y! Please add the exact numbers.
Apologies! Text updated: L568-569: “On average, the bottom depth of the euphotic zone was 14 12 meters deeper than the bottom depth of the mixed layer”
Figure 6: Quite interesting! I’m curious Stations 2 and 9 showed almost the same conversion factors between surface and subsurface. Any possible explanations? Because of intense vertical mixing at both stations?
It is a bit of a befuddling result – Strong similarities between the subsurface and surface samples would make sense in the context of strong upwelling. However, Station 2 (Cape Blanco) and 9 (Cape Mendocino) were very different from one another. Upwelling, estimated from the NOAA CUTI index, was not the same between the two capes (figure below).
Image caption: Study area colored by the NOAA CUTI upwelling index (1km^2 resolution) on May 31 (day of station 2 sampling) and June 10 (day of station 9 sampling). Star shape indicates our position on the day pictured. Indicating Station 2 sampling coincided with strong upwelling while station 9 sampling coincided with strong downwelling. Speculatively, strong vertical mixing can still explain the strong coherence between subsurface and surface samples at these stations although the prevailing direction of vertical transport differed.
Subsections 4.2 and 4.3 are a bit redundant for me even as an aquatic photosynthesis researcher although your discussion is quite interesting! It would be great if you could shorten these paragraphs.
We have re-edited these sections for brevity. While we hope the writing is more concise, we still kept lots of the fundamental information discussed in these sections since they provide critical context for our results.
Subsection 4.3.2: There is no mention of Fe availability. No need here? It would be more interesting if you could include just a few discussions on Fe availability as your manuscript title suggests.
We have updated L993-995: In contrast to , there was no significant differences in NCP:ETRPSII between Cape Blanco, Cape Mendocino, or offshore, suggesting limited effects of nutrient limitation on decoupling between ETRPSII and NCP.
We also added some text that explains our result, per Reviewer 1’s suggestion:
L1035-1047: Despite the inherent dependency of net oxygen production on gross oxygen production, the strength of the correlation between NCP and ETRPSII and the consistency of ETRPSII:NCPacross offshore, Cape Blanco and Cape Mendocino subregions is surprising given the vast suite of potential methodological and physiological sources of uncoupling (Fig 7). However, the derivations of NCP and GP both have similar dependencies on mixed layer Chl concentration. To obtain FRRF-derived GP estimates in comparable units of mmol O2 m-2 d-1, we multiplied in-situ ETRPSII by mixed layer Chl concentrations (Eq 5). While mixed layer Chl concentrations are not explicitly included in NCP calculations (Sect 2.6), biomass is expected to be a primary driver of bulk productivity. If Chl-normalized NCP is instead compared against GP expressed in units of mmol O2 Chl-1 d-1, the correlation between 24h binned and instantaneous NCP and ETRPSII estimates decrease to = 0.22 and 0.35, respectively. We therefore conclude that it remains challenging to derive gross and net carbon fluxes from FRRF measurements alone, but paired ETRPSII and Chl measurements can provide useful constraints for NCP estimates.
L685: OK, I would partially agree with your rough estimation using 400-700 chls/RCII. Is it also applicable to Fe-limited waters such as Cape Mednonico? If you look through other chls/RCII ratios in Fe-limited waters (e.g., Strzepek et al. 2019 with appendices and references therein).
Chl:RCII is expected to increase in Fe limited waters. Previously, Schuback et al. (2015) estimated Chl:RCII as 500 for Fe replete water and 700 for Fe limited water based on literature values from Greene et al., 1992.
We have added significant discussion surrounding Chl:RCII (L889-917) that will be of interest:
“In addition to non-linear electron transport, is also directly affected by the number of Chl energetically coupled to RCII (. Directly measuring requires specialized O2 flash yield instrumentation (Suggett et al., 2009), which was unavailable for this study. However, in-situ Chl concentrations, normalized to FRRF-derived proxies for [RCII] (∝ Fo/) following the approach of Oxborough et al. (2012), can be used to examine variability in between samples. With a known instrument calibration factor, Ka, either provided by instrument manufacturers or determined independently by O2 flash yield measurements, this approximation could be used to estimate the absolute value of .
It is well established that Chl:RCII () ratios increase under low light, to maximize light absorption (Greenbaum and Mauzerall, 1991). In our measurements, the proxy for varied significantly between sample depths, with higher at the bottom of the euphotic zone compared to surface depths, confirmed by a t-test comparison of population means (p << 0.01). Iron limitation is also expected to increase Chl:RCII. Although iron limitation lowers total cellular Chl content, Chl is more likely to be energetically coupled to RCII rather than PSI reaction centers (Greene et al., 1992). Accordingly, displayed a negative correlation with Fea1 expression in surface samples ( = -0.72, p <0.05, n = 9), which we used as a proxy for iron limitation. We thus conclude that Fe-stress likely contributed to variability in in addition to influencing non-linear electron transport. The hyperbolic relationship between carbon fixation and electron transport was unaffected by , which was assumed to be constant for individual samples throughout the course of photosynthesis-irradiance experiments (Appendix 4). However, this assumption may be violated under high light, due to photoinactivation of RCII (Campbell and Serôdio, 2020). A robust understanding of variability requires direct [RCII] measurements collected in parallel with ETRPSII and carbon fixation measurements.”
L731: Energy imbalance due to photosystem impairment can be indeed inferred from NPQNSV as you discuss. I would also suggest looking at qP or qL as described above.
Added in details on qP (referred to in our manuscript as F’q/F’m) per your suggestion (Table 1).
Citation: https://doi.org/10.5194/egusphere-2024-3812-AC2
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-3812', Anonymous Referee #1, 04 Feb 2025
This paper investigates the inter-comparability of 3 different “currency” approaches to measuring primary productivity in the iron-limitation mosaic of the California Current System. In particular, it investigates the influence of iron availability/limitation on the inter-comparability of the approaches. This is useful in that it elucidates adaptations at the cellular scale that allow phytoplankton to cope with iron limitation, and it also shines a light on scenarios that do/do not allow a straightforward conversion from one approach to measuring primary productivity to another. The work is timely, interesting, well-structured and thought through, and the paper is well written, with care clearly taken to allow the reader to follow the arguments and methodologies. The figures are overall good and appropriate, and the main finding, i.e. that a fluorescence-based approach to measuring gross primary productivity is well correlated to daily integrated net community production is indeed surprising and interesting. Below are some very minor comments that hopefully will improve the presentation further.
L90: in the list of respiration pathways for NCP, shouldn’t bacterial respiration as well as respiration by larger animals be added to this list? See also Figure 7.
Figure 1: panels a and b are reversed in the caption.
L151: was PAR measured or assumed? If it was measured, how was it measured?
Eq 4: what is meant by “in situ PAR”? PAR averaged for the MLD? I’m struggling to make sense of the equation.
Eq 7: the superscript “B” – is it a designation (bio?) or an exponent? Can you make it a bit clearer by spelling it out?
L262: inconsistency regarding letter k in the fraction: it’s capitals here, but small-caps above (L253).
L266: since it’s the liquid fraction you’re interested in, should it be “filtered through” rather than “onto”?
Figure 2: could you indicate here what areas were used for the boxplots in Figure 3?
L365-368: These statements could distinguish a bit more between the capes – not all statements are true for both capes (Figure 4).
L378/79: this statement is not true for alpha at Cape Blanco – this whole section could be either a bit more refined, or throw in a few more “overall”s to indicate that what is said is not always true but a general trend.
L380: is beta shown anywhere?
L387: “Elevated values…” – this is only true for Cape Blanco. You said previously that both capes are upwelling, but to different degrees; maybe start early with a clear distinction between the two capes, but “upwelling” is not one of them, that’s true for both (see also L389).
L416/17: Is this the “in situ PAR” that confused me above? If so, make sure to call it that in the Figure.
L425-429: refer to Figure 5 here?
L440-444: maybe add a word here about the uncertainties (due to nitrification in the surface layer) that were mentioned in the methods? What order of magnitude are they?
L451: “x+/-y” wants to be filled with numbers
Figure 7 and its caption: Do you mean “spatial and time scales”? Only the time scales are mentioned in the Figure. Also, I can’t see pathway 7 anywhere in the Figure
L633: Did you mean to refer to Figure 6 here (rather than Figure 8)?
L653: “… with low Cmax” – this should be “…with low Pmax-C”, no?
Figure 8 and its caption: maybe make it clearer in the caption that blue and red also refer to different Fe-limitation regimes, or make a second panel where the dots are coloured by some measure of Fe limitation (Fv/Fm, sigma,… something like that)? This is a useful distinction, no? Also, Vmax (mentioned in the caption) is not shown, nor do I know what it would be based on what has been discussed?
L664: mention of Vmax again – should this be Ksat? It’s getting a bit confusing in this paragraph… if there is both a Vmax and a Ksat, then maybe a table of all the variables and their meaning would be useful?
L696/70: I would have thought that Cyt b559a would correlate with NPQ, and therefore expected a negative correlation with Pmax-C. Does this deserve a bit more discussion?
L678: Vmax again… and also, a question that comes up reading this paragraph and the preceding ones: How do you decide whether to discuss Vmax (Ksat) rather than Pmax-C? Is there a scheme to this? They seem somewhat interchangeable to me at this point. When you don’t mention one or the other, does it mean the correlation would have been insignificant? Again, maybe a word or two on this would be useful.
L678-80: This sentence is either missing a final section, or the “and” after the comma needs to go; also, please specify whether this increased coupling of Chl to RCII is specific to iron limitation or generally true.
L681: “increases in”? or “increased”? This whole paragraph could be a bit clearer - what does the literature say, what was measured here, what needs to be assumed, how does it all come together?
L704: Can you insert the actual ranges that have been observed by others?
Paragraph starting L698: The observed relationship between FRRf-derived GPP and NCP is a significant and surprising result (maybe you disagree, but then please lay out more clearly why it’s not surprising). Maybe a bit more discussion on why this relationship with NCP works here would not be amiss. Where/when do we expect this to work elsewhere (and where/when may it fall over)? What needs to be investigated next? Does it have anything to do with iron at all?
Figure 9 and caption: Does the correlation for panel b (rho = 0,92,L715) include the negative data point? How should one think about the negative data point, can it be explained, could it have to do with upwelling/downwelling? And what about the intercept for the line of best fit (-0.55), how to think about that? That it’s very close to zero and that’s good, or does it have some other meaning that’s worth spelling out? Also, the final sentence of the figure caption is a repeat, except for the r2. Regarding panel A of the figure: could it be improved by plotting the points as a heat map, to show where most of the data points fall?
L739: Regarding “fine-scale variability”: what scales are being resolved when we go to 24h binning?
Citation: https://doi.org/10.5194/egusphere-2024-3812-RC1 -
AC1: 'Reply on RC1', Yayla Sezginer, 26 Mar 2025
We thank reviewer 1 for their positive feedback and edits, which strengthen this study. Reviewer 1's suggestions are in bold and our responses are listed below. All edits are referenced by line number to the corresponding file of manuscript track changes.
L90: in the list of respiration pathways for NCP, shouldn’t bacterial respiration as well as respiration by larger animals be added to this list? See also Figure 7.
We have amended the statement to ‘community wide’ respiration to cover respiration by respiration by bacteria and macrofauna. L110 in track changes file.
Figure 1: panels a and b are reversed in the caption.
Corrected, L143-144.
L151: was PAR measured or assumed? If it was measured, how was it measured?
Here, PAR is delivered by instrument LED lamps that provide continuous actinic light. PAR provided by the instrument was not directly measured during ETRPSII measurements, but the instrument LEDs were calibrated against a WALZ ULM-500 PAR meter prior to deployment. Note on calibration added L180
Eq 4: what is meant by “in situ PAR”? PAR averaged for the MLD? I’m struggling to make sense of the equation.
Yes, PARin-situ refers to the mean PAR in the mixed layer. This equation closely follows that Eq. 3 in Domingues and Barbosa (2023), although some of the parameter names differ between our manuscript and their paper to maintain consistency within our manuscript. For example, here we denote the mixed layer depth as MLD, while Dominges and Barbosa use ‘Zm’. Some text added for clarification (L214).
Eq 7: the superscript “B” – is it a designation (bio?) or an exponent? Can you make it a bit clearer by spelling it out?
Yes, B for Bio. Clarified text reads as follows: “Biological concentrations, indicated by the superscript, ‘B’, are derived by isolating and removing physical solubility effects from measured gas concentrations.” (L316).
L262: inconsistency regarding letter k in the fraction: it’s capitals here, but small-caps above (L253).
Corrected (L329)
L266: since it’s the liquid fraction you’re interested in, should it be “filtered through” rather than “onto”?
Corrected (L361).
Figure 2: could you indicate here what areas were used for the boxplots in Figure 3?
New panel added to Figure 3 to indicate data allocation between subregions.
Note that although some ‘Cape Blanco’ data between 42.5-41.5N are at a relatively western longitude they are designated part of the coastal upwelling plume based on their low SST. Subregions were identified according to latitude and temperature criteria. Samples were considered part of an upwelling group if they had SST < 12 oC. If upwelling samples were north of 41N they were considered part of the Cape Blanco plume. If south of 41N, they were designated Cape Mendocino samples. All other samples were grouped in the offshore category. Criteria were determined by studying Figure 1a.
L365-368: These statements could distinguish a bit more between the capes – not all statements are true for both capes (Figure 4).
In reviewing Figure 4, there is evidence of diel cycles in Fv/Fm, , and NPQNSV. However, it is true that the magnitude of apparent diel cycles varies between Cape Blanco and Cape Mendocino. To acknowledge that, we have added the statement, ‘the magnitude of diel variability in FV/FM, , and NPQNSV signals displayed significant variability between subregions, as discussed below’ (L489). In-depth analysis of differences in photo-physiology between the two capes is discussed in the following paragraphs and Table 1.
Additionally, there is some convolution of the diel signals since the ship typically stayed close to the shelf during the day and transited offshore overnight. We have already acknowledged this in 507-514; ‘We note, however, that there is potential for some convolution of temporal and spatial variability, as the ship spent more time offshore in the night, and on-shore during the daytime. It is thus possible, that some of the diel cycling partially reflects different photo-physiological signals between coastal and offshore waters.’
L378/79: this statement is not true for alpha at Cape Blanco – this whole section could be either a bit more refined, or throw in a few more “overall”s to indicate that what is said is not always true but a general trend.
Threw in some generalizing terms: ‘Generally’ (L484), ‘Overall’ (L501)
L380: is beta shown anywhere?
Beta is included in Table 1 now.
L387: “Elevated values…” – this is only true for Cape Blanco. You said previously that both capes are upwelling, but to different degrees; maybe start early with a clear distinction between the two capes, but “upwelling” is not one of them, that’s true for both (see also L389).
Updated the text to reflect the general trends and relationships reported in Table 1 between upwelling conditions and photo-physiology, while still setting up the reader to understand how/why photo-physiology can still differs between two upwelling areas.
L516-523:
In general, FV/FM, Pmax, Ek and displayed positive relationships with upwelling indicators, i.e. salinity, nitrate, and decreased sea surface temperature (Table 1), suggesting that vertical transport of nutrient-rich water to the surface supported high photochemical yields. In contrast, signs of upwelling were associated with decreased and NPQNSV. However, despite general trends between photo-physiological parameters and upwelling, there were significant differences in photo-physiological properties between the Cape Blanco and Cape Mendocino upwelling plumes. At Cape Blanco…
L416/17: Is this the “in situ PAR” that confused me above? If so, make sure to call it that in the Figure.
Yes, added to figure 5a legend and caption.
L425-429: refer to Figure 5 here?
Done (L579).
L440-444: maybe add a word here about the uncertainties (due to nitrification in the surface layer) that were mentioned in the methods? What order of magnitude are they?
Added an estimate of error due to euphotic zone nitrification and added some details to the methods describing how this error was determined.
L331-341:
“We note that several recent studies have observed nitrification within the euphotic zone, challenging the assumption that N2O production is limited to subsurface waters (Grundle, Juniper and Giesbrecht, 2013; Smith et al., 2014), and potentially leading to overestimates in our vertical mixing-corrected NCP estimates. Previous observations in the CCS reported a range of depth-integrated mixed layer nitrification rates between 0.3 – 2 mmol NH4+ m-2 d-1, resulting in consumption of 0.6 – 4 mmol O2 m-2 d-1 (Stephens et al., 2020). Following the approach of Izett et al. (2018) we used a range of realistic N2O:O2 stoichiometries to estimate potential upper and lower bounds of mixed layer N2O production. We determined mixed layer N2O production likely ranged between 0.09 – 0.23 mol N2O m-2 d-1, which would yield offsets in our final NCP estimates between 1.2 and 3.3 mmol O2 m-2 d-1. Total uncertainty due to sources of error in other derived parameters was determined by following Izett (2021).
L594 – 596:
“Uncertainty in vertical-mixing corrected NCP due potential mixed layer nitrification (see sect. 2.6) represented between 1.5 – 4.2% of our mean corrected NCP value.”
L451: “x+/-y” wants to be filled with numbers
Yikes! Amended with the values.
L608-609: “On average, the bottom depth of the euphotic zone was 14 12 meters deeper than the bottom depth of the mixed layer”
Figure 7 and its caption: Do you mean “spatial and time scales”? Only the time scales are mentioned in the Figure. Also, I can’t see pathway 7 anywhere in the Figure
Removed ‘spatial’ and added a 7 symbol in the NCP panel.
L633: Did you mean to refer to Figure 6 here (rather than Figure 8)?
Removed the reference to figure 8, which does not belong here. L791
L653: “… with low Cmax” – this should be “…with low Pmax-C”, no?
Corrected.
Figure 8 and its caption: maybe make it clearer in the caption that blue and red also refer to different Fe-limitation regimes, or make a second panel where the dots are coloured by some measure of Fe limitation (Fv/Fm, sigma,… something like that)? This is a useful distinction, no? Also, Vmax (mentioned in the caption) is not shown, nor do I know what it would be based on what has been discussed?
Vmax in caption corrected to Pmax-C. Added some text to the Figure 8 caption.
L664: mention of Vmax again – should this be Ksat? It’s getting a bit confusing in this paragraph… if there is both a Vmax and a Ksat, then maybe a table of all the variables and their meaning would be useful?
Apologies for the confusion between Vmax, Pmax-C and Cmax. Pmax-C is the maximum Chl-normalized carbon fixation rate. The parameter pmax-C is comparable with Vmax in traditional Micahelis-Menten terminology. However, the maximum carbon fixation rate, whether expressed as a function of light, as in a PI curve, or as a function of ETR is nearly the same (no difference in the maximum measured carbon fixation rate, but the derived value changes very slightly between models). When writing this manuscript, there were many iterations where we tested different parameter names for Pmax-C, and some of the old names, Cmax, Vmax, etc, were not caught during editing. Thank you for your careful attention to this detail! Section 4.3.1 has been carefully reviewed now to catch all mentions of Cmax/Vmax/or Pmax-C and make sure they all read Pmax-C.
L696/70: I would have thought that Cyt b559a would correlate with NPQ, and therefore expected a negative correlation with Pmax-C. Does this deserve a bit more discussion?
I would agree that cyt b559a would be expected to positively correlate with NPQ, however I would disagree that Pmax should be negatively correlated with NPQ. Samples with high and low Pmax-C both expressed high levels of NPQ. One of the main findings of this section is that in some cases, carbon fixation cannot be predicted from NPQ. The text has been modified as follows:
While the precise functional roles of Cyt b559a are still not certain, previous studies have demonstrated its potential role in photoprotective cyclic electron transport around PSII and PSII assembly (Chiu and Chu, 2022). L952-954.
L678: Vmax again… and also, a question that comes up reading this paragraph and the preceding ones: How do you decide whether to discuss Vmax (Ksat) rather than Pmax-C? Is there a scheme to this? They seem somewhat interchangeable to me at this point. When you don’t mention one or the other, does it mean the correlation would have been insignificant? Again, maybe a word or two on this would be useful.
Text around here has changed quite a bit so that text is gone now (see following comment). However, the point about interchangeable parameters is still a good one.
The text has been cleaned up so only Pmax-C is mentioned throughout. Kssat and Pmax-C are not interchangeable. Rather, the ksat and Pmax-C terms discussed here are comparable with the Ksat and Vmax terms in a Michealis-Menten equation.
Discussion is mostly limited to Pmax-C for simplicity and because this parameter is more representative of photosynthetic enzymes concentrations and kinetics, which provides mechanistic insights into photosynthetic functioning and are useful to understand the observed variability in e:C/. It’s worth noting that Kssat and Pmax-C are positively correlated ( = 0.70, p << 0.01) L954-956.
L678-80: This sentence is either missing a final section, or the “and” after the comma needs to go; also, please specify whether this increased coupling of Chl to RCII is specific to iron limitation or generally true.
Lots of new details added to the discussion of Chl:RCII and its influence on our comparison of carbon fixation and electron transport rates:
“In addition to non-linear electron transport, is also directly affected by the number of Chl energetically coupled to RCII (. Directly measuring requires specialized O2 flash yield instrumentation (Suggett et al., 2009), which was unavailable for this study. However, in-situ Chl concentrations, normalized to FRRF-derived proxies for [RCII] (∝ Fo/) following the approach of Oxborough et al. (2012), can be used to examine variability in between samples. With a known instrument calibration factor, Ka, either provided by instrument manufacturers or determined independently by O2 flash yield measurements, this approximation could be used to estimate the absolute value of .
It is well established that Chl:RCII () ratios increase under low light, to maximize light absorption (Greenbaum and Mauzerall, 1991). In our measurements, the proxy for varied significantly between sample depths, with higher at the bottom of the euphotic zone compared to surface depths, confirmed by a t-test comparison of population means (p << 0.01). Iron limitation is also expected to increase Chl:RCII. Although iron limitation lowers total cellular Chl content, Chl is more likely to be energetically coupled to RCII rather than PSI reaction centers (Greene et al., 1992). Accordingly, displayed a negative correlation with Fea1 expression in surface samples ( = -0.72, p <0.05, n = 9), which we used as a proxy for iron limitation. We thus conclude that Fe-stress likely contributed to variability in in addition to influencing non-linear electron transport. The hyperbolic relationship between carbon fixation and electron transport was unaffected by , which was assumed to be constant for individual samples throughout the course of photosynthesis-irradiance experiments (Appendix 4). However, this assumption may be violated under high light, due to photoinactivation of RCII (Campbell and Serôdio, 2020). A robust understanding of variability requires direct [RCII] measurements collected in parallel with ETRPSII and carbon fixation measurements.”
L681: “increases in”? or “increased”? This whole paragraph could be a bit clearer - what does the literature say, what was measured here, what needs to be assumed, how does it all come together?
This text has been replaced with the paragraphs above.
L704: Can you insert the actual ranges that have been observed by others?
Kranz et al. (2020) observed GPP rates of 0 – 4000 mmol C m-2 d-1, 0 – 4000 mmol O2 m-2 d-1 from O2/Ar- and FRRF- based approaches. This range is coincident with that observed in this study (L1047).
Paragraph starting L698: The observed relationship between FRRf-derived GPP and NCP is a significant and surprising result (maybe you disagree, but then please lay out more clearly why it’s not surprising). Maybe a bit more discussion on why this relationship with NCP works here would not be amiss. Where/when do we expect this to work elsewhere (and where/when may it fall over)? What needs to be investigated next? Does it have anything to do with iron at all?
Agreed! This is a significant and perhaps surprising result, due to the large differences in integration time scales of GPP and NCP, and the potential for different metabolic and ecological processes to decouple these rates (L918).
We initially hypothesized that ETR:NCP would vary with ETR:C-fixation, and that factors, like Fe availability, that drive decoupling between ETR and GPP would similarly decouple ETR and NCP. However, we did not observe significant differences in ETR:NCP between Cape Blanco and Cape Mendocino as would be expected if Fe was playing a key role in ETR:NCP variability. Rather, the relationship between ETR:NCP was fairly consistent across the study region. We offer an alternative explanation for the surprising consistency in the ETR:NCP relationship below (L1093-1113):
Despite the inherent dependency of net oxygen production on photosynthetic electron transport rates, the strong correlation between NCP and ETRPSII and the consistency of ETRPSII:NCPacross our entire study regions is surprising given the large number of methodological and physiological factors that can significantly uncouple these rates (Fig 7). However, NCP and GP estimates both have similar dependencies on mixed layer Chl concentration. To obtain FRRF-derived GP estimates in comparable units of mmol O2 m-2 d-1, we multiplied in-situ ETRPSII by mixed layer Chl concentrations (Eq 5). Although mixed layer Chl concentrations are not explicitly included in NCP calculations (Sect 2.6), biomass is expected to be a primary driver of bulk productivity. Indeed, when we compared Chl-normalized ETRPSII and NCP estimates, we found a much weaker relationship between with correlation coefficients decreasing to 0.22 and 0.35 for 24h binned and instantaneous measurements, respectively. We therefore conclude that it remains challenging to derive gross and net carbon fluxes from FRRF measurements alone, but paired ETRPSII and Chl measurements can provide useful constraints for NCP estimates.
Figure 9 and caption: Does the correlation for panel b (rho = 0,92,L715) include the negative data point? How should one think about the negative data point, can it be explained, could it have to do with upwelling/downwelling? And what about the intercept for the line of best fit (-0.55), how to think about that? That it’s very close to zero and that’s good, or does it have some other meaning that’s worth spelling out? Also, the final sentence of the figure caption is a repeat, except for the r2. Regarding panel A of the figure: could it be improved by plotting the points as a heat map, to show where most of the data points fall?
The correlation does include the negative data point. This point corresponds to measurements on June 2. On this day NCP reached its minima. This point corresponded with high SST indicative of offshore waters outside the upwelling plume (see Fig 5b).
The y-intercept being close to zero is a positive indication that our 24h binned NCP and ETR comparison are reasonable, however this y-intercept should not be taken as an absolute. As we can see from the highly negative point in Fig 9b, very negative NCP values are possible in net heterotrophic waters, while ETR values can never be net negative.
Repeat sentence removed from Fig 9 caption
A heat map of the data density indicates that most of the data is concentrated around [ETR = 0, ncp = 100]. This makes sense since ~half the ETR data should = 0 overnight and the mean ncp was about 80 mmol O2 m-2 s-1. Measurements collected during the day (ETR > 0) appear evenly spread. Since the point that ETR must equal 0 overnight is made a few times in the text, we have decided not to change the manuscript in the figure to the heat map, but hope it is of interest for Reviewer 1.
L739: Regarding “fine-scale variability”: what scales are being resolved when we go to 24h binning?
This is an excellent point! A primary motivation for using FRRF and O2/N2 is the ability to capture fine-scale variability. That fine-scale variability is of course smeared if a 24h bin is applied. The spatial ‘smear’ will depend on the boat’s movement throughout the day. In our case, the boat stayed on-station during the day for full-scale morning and afternoon sampling programs and transited to the following station overnight.
Here each 24h binned ETR measurement roughly represents the mean photochemical activity at each station over 24 hours, since ETR is zero overnight when the ship was transiting. The interpretation is more complicated for NCP which captures variability in O2/N2’ overnight due to variable respiration rates as the ship transited through different water masses.
Binning obviously negates the critical advantages of high frequency measurement systems like the FRRF and PIGI. However, we applied this approach to purposefully smear the temporal resolution of ETR to approximate the time-scales represented by O2/N2’-based NCP measurements and provide a more just comparison. This approach is far from perfect, but we found it useful to illustrate how trying to account for differences in time-scales can improve the correlation between productivity metrics.
Citation: https://doi.org/10.5194/egusphere-2024-3812-AC1
-
AC1: 'Reply on RC1', Yayla Sezginer, 26 Mar 2025
-
RC2: 'Comment on egusphere-2024-3812', Kazuhiro Yoshida, 28 Feb 2025
Review Report
Sezginer et al. estimated primary productivity in two different upwelling areas with multiple approaches for photosynthesis measurements. It has been controversial but also dreamt estimation of primary production by measuring real-time and/or autonomous chl fluorescence measurements; however, few studies have addressed the discrepancies between fluorescence and carbon fixation/oxygen evolution particularly in natural waters. The electron-carbon conversions were successful in restricted and homogenous waters (e.g., Cheah et al. 2011). Sezgniger et al. comprehensively compared FRRf, O2, and carbon on a large scale within different Fe-mosaiced upwelling systems. They did not just measure primary production; they also mentioned underlying transcriptional changes along with different photosynthetic performance and Fe availability. Overall, they clarify the hope and limitations of the dream bridge between electron-oxygen-carbon. Although the manuscript reads well, I suggest further analysis or clarifications of your interpretation of the data. I thus recommend Moderate revision before publication.
[Suggestions]
A few mentions of phytoplankton taxonomy
- During your sampling, diatoms, particularly Cheatoceros, dominated the phytoplankton communities in both upwelling areas. Undoubtedly, diatoms were most abundant in your samples; however, you might also observe other diatoms, dinoflagellates and/or small taxa. You conducted light microscopy to look at the phytoplankton taxonomy of your samples. If your microscopy data was sufficiently reliable, subsection 4.2 can be improved as Taguchi (1976), Finkel (2000), and Suggett et al. (2009) reported that different taxa showed unique photosynthetic performances even both carbon and fluorescence.
More discussion on Carbon/nitrogen ratio
- I agree Fe availability modifies the photosynthesis machinery such as PSI: PSII and Cytochrome: PSII, which decouples the electron: carbon conversion. However, photosynthetic energy can be allocated to both carbon and nitrogen et cetera, which could also decouple the conversion factor. You only discuss C: N ratios compared with the Redfield ratio. Discussion on the relative allocation to nitrogen from C: N ratios can improve your discussion on the decoupling electron: carbon.
Derivation of qP or qL
Your discussion on low Fe effects on FRRf and carbon fixation was on non-linear electron transport due to an imbalance between photosystems. I understand you well discuss the reduction of PSI, inferred from TPMs (or relative gene expression) of the PSI-related genes. In addition, I would suggest you calculate 1-qP or 1-qL as the state of open/closed PSII from your FRRf data.
How much RNA data did you throw away?
I agree your transcriptomics indeed deepened your discussion. I have two technical questions:
(1) Have you compared with other de novo assemblers than rnaSpades or CD-HIT-EST? Any assembling performance data? Did you pre-process raw RNA sequences?
- (2) You extracted diatom reads/contigs for further downstream analysis. However, transcriptomic data might be unbiased and might have various reads/contigs from various organisms. What proportion did diatom data contribute to the total sequences? How much RNA data did you throw when only focusing on diatoms?
[Minor issues]
L110: Have you validated your nitrate sensor with your autoanalyzer outputs?
L139: How did you dissimilate/separate QA and QB reoxidation from your FRRf data? You can separate QA and QB if using an algal isolate, but was it appropriately separated in natural communities even with various Fe availabilities?
L203: I understand NPQ fitting is quite tricky. Serôio and Lavaud (2011) thus discussed model performance on NPQ-E fittings. How were your single-component exponential fittings? Any stats outputs?
L206: Why did you choose 46% light depth? You surely collected the same light depths at each station? Or simply light intensities at ~50 m almost similar at each station?
L250: Any other source(s) of N2O other than photoinhibition?
L321 and elsewhere: How did you measure salinity? If from CTD, psu is no longer used. Just unitless.
L354: In general, the Redfield ratio does not apply to surface waters I reckon because Redfield ratio 16 was measured in deep waters. I would rather recommend C: N allocation ratios in algal cells as discussed above.
Figure 7: As you discuss, Cape Bianco had lower temperatures due to stronger upwelling. It is quite interesting but did this temperature difference affect (1) the enzymatic activity of xanthophyll pigment synthesis and consequently NPQ dynamics/xanthophyll cycle (XC)? and (2) transcriptomic regulation such as low-temperature adaptation, which overlayed Fe-regulated gene expression or TPM counts?
Regarding (1), NPQ correlated with SST in Table 1, which might be due to temperature-dependence XC or NPQ? Also, differences in the community structure might influence NPQ or other FRRf parameters?
Table 1: In the 3rd row, Pmax should be ETRmax??
Figure 5(c): GP at Cape Bianco is highly variable with numerous spiky up-downs compared with the peaceful Cape Mendocino. Any possible explanations for the difference? Or simply, large errors for estimation of GP in Cape Bianco?
L451: I wanna know x and y! Please add the exact numbers.
Figure 6: Quite interesting! I’m curious Stations 2 and 9 showed almost the same conversion factors between surface and subsurface. Any possible explanations? Because of intense vertical mixing at both stations?
Subsections 4.2 and 4.3 are a bit redundant for me even as an aquatic photosynthesis researcher although your discussion is quite interesting! It would be great if you could shorten these paragraphs.
Subsection 4.3.2: There is no mention of Fe availability. No need here? It would be more interesting if you could include just a few discussions on Fe availability as your manuscript title suggests.
L685: OK, I would partially agree with your rough estimation using 400-700 chls/RCII. Is it also applicable to Fe-limited waters such as Cape Mednonico? If you look through other chls/RCII ratios in Fe-limited waters (e.g., Strzepek et al. 2019 with appendices and references therein).
L731: Energy imbalance due to photosystem impairment can be indeed inferred from NPQNSV as you discuss. I would also suggest looking at qP or qL as described above.
Review Report
Sezginer et al. estimated primary productivity in two different upwelling areas using multiple approaches for photosynthesis measurements. The estimation of primary production by measuring real-time and/or autonomous chl fluorescence has been controversial but also considered a dream; however, few studies have addressed the discrepancies between fluorescence and carbon fixation/oxygen evolution, particularly in natural waters. Electron-carbon conversions have been successful in restricted and homogeneous waters (e.g., Cheah et al. 2011). Sezginer et al. comprehensively compared FRRf, O₂, and carbon on a large scale within different Fe-mosaiced upwelling systems. They did not only measure primary production; they also examined underlying transcriptional changes along with different photosynthetic performances and Fe availability. Overall, they clarify both hopes and limitations of the dream bridging between electron-oxygen-carbon. Although the manuscript reads well, I suggest further analysis or clarifications of your data interpretation. I, therefore, recommend moderate revision before publication.
[Suggestions]
- A few mentions of phytoplankton taxonomy
During your sampling, diatoms, particularly Chaetoceros, dominated the phytoplankton communities in both upwelling areas. Undoubtedly, diatoms were the most abundant taxa in your samples; however, you might have also observed other diatoms, dinoflagellates, and/or small taxa. You conducted light microscopy to examine the phytoplankton taxonomy of your samples. If your microscopy data were sufficiently reliable, subsection 4.2 could be improved, as Taguchi (1976), Finkel (2000), and Suggett et al. (2009) reported that different taxa exhibit unique photosynthetic performances in both carbon fixation and fluorescence. - More discussion on Carbon/Nitrogen ratio
I agree that Fe availability modifies the photosynthesis machinery, such as PSI:PSII and Cytochrome:PSII, which decouples electron:carbon conversion. However, photosynthetic energy can be allocated to both carbon and nitrogen, among other processes, which could also decouple the conversion factor. You only discuss C:N ratios compared with the Redfield ratio. A discussion on the relative allocation to nitrogen based on C:N ratios could improve your analysis of the decoupling between electron and carbon conversion. - Derivation of qP or qL
Your discussion on the effects of low Fe on FRRf and carbon fixation focuses on non-linear electron transport due to an imbalance between photosystems. While you provide a well-developed discussion on PSI reduction, inferred from TPMs (or relative gene expression) of PSI-related genes, I suggest calculating 1-qP or 1-qL to assess the state of open/closed PSII from your FRRf data.
- How much RNA data did you discard?
I agree that your transcriptomic analysis significantly deepened the discussion. However, I have two technical questions:
(1) Have you compared other de novo assemblers besides rnaSpades or CD-HIT-EST? Do you have any assembly performance data? Did you pre-process raw RNA sequences?
(2) You extracted diatom reads/contigs for further downstream analysis. However, transcriptomic data are typically unbiased and may include reads/contigs from various organisms. What proportion of the total sequences was attributed to diatoms? How much RNA data did you discard when focusing only on diatoms?
[Minor issues]
L110: Have you validated your nitrate sensor against your autoanalyzer outputs?
L139: How did you distinguish/separate QA and QB reoxidation from your FRRf data? While separation is possible with an algal isolate, was it appropriately achieved in natural communities with varying Fe availabilities?
L203: I understand that NPQ fitting is quite tricky. Serôdio and Lavaud (2011) discussed model performance on NPQ-E fittings. How well did your single-component exponential fittings perform? Any statistical outputs?
L206: Why did you choose 46% light depth? Did you collect samples at the same light depth for each station, or were light intensities at ~50 m similar across stations?
L250: Were there any additional sources of N₂O other than photoinhibition?
L321 and elsewhere: How did you measure salinity? If you used a CTD, note that PSU is no longer used and should be reported as a unitless value.
L354: The Redfield ratio generally does not apply to surface waters, as it was derived from deep-water measurements. I suggest discussing C:N allocation ratios in algal cells instead.
Figure 7: As you noted, Cape Bianco had lower temperatures due to stronger upwelling. Did this temperature difference influence (1) the enzymatic activity related to xanthophyll pigment synthesis and, consequently, NPQ dynamics/xanthophyll cycling (XC), or (2) transcriptomic regulation, such as low-temperature adaptation, which could have overlaid Fe-regulated gene expression or TPM counts? Regarding (1), NPQ correlates with SST in Table 1. Could this be due to temperature-dependent XC or NPQ? Additionally, could differences in community structure have influenced NPQ or other FRRf parameters?
Table 1: In the third row, should Pmax be labeled ETRmax instead?
Figure 5(c): GP at Cape Bianco is highly variable, with numerous spiky ups-downs compared to the stable and peaceful Cape Mendocino. What might explain this difference? Or is it simply due to large estimation errors for GP at Cape Bianco?
L451: I wanna know x and y! Please provide the exact values for x and y.
Figure 6: Very interesting! Stations 2 and 9 showed nearly identical conversion factors between surface and subsurface. Any possible explanations? Could intense vertical mixing at both stations be responsible?
Subsections 4.2 and 4.3: These sections feel somewhat redundant, even for an aquatic photosynthesis researcher like myself. While your discussion is quite interesting, consider shortening these paragraphs for clarity.
Subsection 4.3.2: There is no mention of Fe availability here. Is this intentional? Including a brief discussion on Fe availability would make this section more aligned with your manuscript’s title.
L685: I partially agree with your rough estimation using 400-700 chls/RCII. However, is this also applicable to Fe-limited waters such as Cape Mendocino? I recommend reviewing other chl/RCII ratios in Fe-limited waters (e.g., Strzepek et al. 2019, including appendices and references therein).
L731: As you discuss, energy imbalance due to photosystem impairment can be inferred from NPQNSV. Additionally, I suggest considering qP or qL, as described above.
Citation: https://doi.org/10.5194/egusphere-2024-3812-RC2 -
AC2: 'Reply on RC2', Yayla Sezginer, 26 Mar 2025
Publisher’s note: this comment was edited on 7 April 2025. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.
We similarly thank Reviewer 2 for their positive feedback, expert insights, and suggestions for clarification and deeper analysis. We have aimed to incorporate as many of their suggestions as possible. Unfortunately, not all the datasets they have requested for deeper analysis are available. Where further data analysis is not possible, we strive to incorporate the additional considerations Reviewer 2 suggests in our discussion. Again, Reviewer 2’s suggestions are listed below in bold with our beneath. Line references in our replies correspond to the lines in the updated manuscript with track changes.
A few mentions of phytoplankton taxonomy
- During your sampling, diatoms, particularly Cheatoceros, dominated the phytoplankton communities in both upwelling areas. Undoubtedly, diatoms were most abundant in your samples; however, you might also observe other diatoms, dinoflagellates and/or small taxa. You conducted light microscopy to look at the phytoplankton taxonomy of your samples. If your microscopy data was sufficiently reliable, subsection 4.2 can be improved as Taguchi (1976), Finkel (2000), and Suggett et al. (2009) reported that different taxa showed unique photosynthetic performances even both carbon and fluorescence.
Our taxonomic analysis primarily relied on chemotaxonomic analysis of pigment data which does not differentiate between diatom taxa. Microscopic analysis was limited to surface samples and was useful to validate our chemotaxonomic analysis and identify general trends between stations, i.e. presence of Pseudo-Nitszchia at Cape Mendocino, but not Cape Blanco. However, this analysis is pretty biased towards larger more recognizable cells and was used more qualitatively. This data is likely not sufficient to apply robust statistical analyses to evaluate taxonomically driven variability in photo-physiology. This is unfortunate since, as you say, it is clear from the literature that different taxa express different optical phenotypes influencing Chlorophyll fluorescence signatures. In a field study, you could argue that variability in these optical phenotypes still ultimately derives from the environmental conditions that select for different phytoplankton taxa.
More discussion on Carbon/nitrogen ratio
- I agree Fe availability modifies the photosynthesis machinery such as PSI: PSII and Cytochrome: PSII, which decouples the electron: carbon conversion. However, photosynthetic energy can be allocated to both carbon and nitrogen et cetera, which could also decouple the conversion factor. You only discuss C: N ratios compared with the Redfield ratio. Discussion on the relative allocation to nitrogen from C: N ratios can improve your discussion on the decoupling electron: carbon.
Agreed, N-assimilation is a decoupling pathway that is not directly triggered by excess energy and could therefore help explain why NPQ and e:C are not always correlated. Further, the source of N is also likely to affect e:C as growth on more reduced forms (i.e. NH4) will divert less e from carbon fixation. However, N-assimilation is also negatively impacted by Fe-limitation as both nitrate and nitrite reductase use Fe as cofactors. Allen et al. (2008) demonstrated these enzymes were downregulated under Fe limitation. While previous studies of diatoms have found either no change in cellular N:P or decreases in N:P in response to Fe limitation (Price 2005; La Roche et al., 2003). As a result, it is not clear how C:N may have varied between our Fe rich and Fe stressed study sites, or how Fe limitation may have influenced e:C decoupling through nitrate assimilation. This is an important point to resolve for the community moving forward. Unfortunately, without the accompanying data, we feel this topic is somewhat out of the scope of our paper.
Derivation of qP or qL
Your discussion on low Fe effects on FRRf and carbon fixation was on non-linear electron transport due to an imbalance between photosystems. I understand you well discuss the reduction of PSI, inferred from TPMs (or relative gene expression) of the PSI-related genes. In addition, I would suggest you calculate 1-qP or 1-qL as the state of open/closed PSII from your FRRf data.
Yes, qP (fraction of open RCII) plays an important role in determining ETRPSII, and is actually included in our ETRPSII equation (Eq.1) where it is shown under its other notation, F’q/F’v (see Tortell, Suggett and Schuback 2023 for a list of synonymous FRRF nomenclature). While we did discuss potential Fe effects on F’q/F’v (L752-758), we did not take the next steps in actually analyzing our F’q/F’v data. Following Reviewer 2’s suggestion, we have now actually examined qP along the cruise transect (see below).
The figure above displays the parameter 1-qP. qP was measured at light levels ranging from 0-850 uE during photosynthesis-irradiance curves. To estimate the fraction of closed RCII, in-situ, we applied the same approach we used to determine underway NPQ, where 1-qP was plotted against PAR, and fit with an exponential curve (92% of model fits had R2 > 0.9). In-situ 1-qP was then estimated by inputting in-situ PAR into the resulting model equation. Our results demonstrate strong diel patterns due to the immediate dependence of this approach on PAR, but also highlight that the fraction of RCII closure during peak midday irradiance was typically lower for Cape Blanco samples (5/30 – 06/04) compared to Cape Mendocino (06/06 – 06/10), with the exception of data collected on 06/02, which is when we transited offshore (note, this is also where we recorded our lowest NCP measurement).
To minimize confusion and avoid adding a new parameter name in the text, we are opting not to use the qP terminology and remain consistent with F’q/F’v. We have added analysis of F’q/F’v to Table 1, and added a note to our discussion section (L770-772):
‘Indeed, measured during underway Photosynthesis-Irradiance curves and mapped onto in-situ irradiances, demonstrated that was higher around Cape Blanco compared to Cape Mendocino (Table 1)’
We also added some text to the methods (Section 2.4) to explain how we evaluated in-situ F’q/F’v. (L256-259)
How much RNA data did you throw away?
Please see the table below contributed by coauthor Emily Speciale detailing our sequencing stats.
Station
Rep
Total Reads
# Reads Mapped
% Reads Mapped
# Reads Taxonomically Annotated
% Reads Taxonomically Annotated
# Protist Group
% Protist Group
# Diatom
% Diatom
1
A
20919052
15566516
74.41
12163291
58.14
8529884
40.78
4462784
21.33
B
20605487
14828573
71.96
11629423
56.44
8495042
41.23
5035031
24.44
2
A
20370660
16757515
82.26
12441421
61.08
9060238
44.48
3337286
16.38
B
33411917
23851899
71.39
17576726
52.61
3878516
11.61
1577339
4.72
C
20451261
16318259
79.79
13878818
67.86
6014888
29.41
3218519
15.74
3
A
31854033
24637176
77.34
17205795
54.01
8931444
28.04
1585515
4.98
B
23102631
18257440
79.03
12937513
56.00
7632094
33.04
1385720
6.00
C
22184188
9543967
43.02
5778682
26.05
1516238
6.83
601494
2.71
4
A
31244147
22594019
72.31
14833206
47.48
8988587
28.77
3345948
10.71
B
17121856
13681984
79.91
9435876
55.11
6813257
39.79
2557063
14.93
C
21784053
16405679
75.31
12470858
57.25
9922004
45.55
5794848
26.60
5
A
24551450
19993489
81.44
16862370
68.68
15686918
63.89
3354493
13.66
B
20633218
15952493
77.31
13714182
66.47
12762334
61.85
3482216
16.88
C
13037287
5936298
45.53
4437254
34.04
632351
4.85
253280
1.94
6
A
18605896
14369514
77.23
11475963
61.68
8299338
44.61
2787489
14.98
B
20757774
7441701
35.85
3668990
17.68
1359173
6.55
441358
2.13
C
22992753
18563929
80.74
14898492
64.80
10666864
46.39
4200265
18.27
7
A
13046360
10221886
78.35
6465869
49.56
4729515
36.25
819835
6.28
B
22985192
18162597
79.02
13223369
57.53
8650540
37.64
1678663
7.30
C
17752669
8665680
48.81
4882063
27.50
2656758
14.97
840107
4.73
8
A
28417725
21660636
76.22
12402752
43.64
8187261
28.81
1294051
4.55
B
18902138
15344332
81.18
9352687
49.48
5208950
27.56
901508
4.77
C
18139851
12084503
66.62
9296668
51.25
6835677
37.68
1183001
6.52
9
A
22252339
17619164
79.18
14355113
64.51
13217881
59.40
1699119
7.64
B
22274606
17295073
77.64
14039185
63.03
12776475
57.36
1600294
7.18
C
18592697
15045958
80.92
11457739
61.62
10642495
57.24
1281205
6.89
10
A
19985953
17084390
85.48
11958397
59.83
8132358
40.69
1241650
6.21
B
14932416
12288932
82.30
8724712
58.43
5509449
36.90
791170
5.30
C
19818129
11020988
55.61
6792417
34.27
2580105
13.02
628061
3.17
SUM
620727738.00
451194590.00
72.69
328359831.00
52.90
218316629.60
35.17
61379310.90
9.89
AVG
21404404.76
15558434.14
72.28
11322752.79
52.62
7528159.64
35.35
2116527.96
9.90
STD. DEV.
4905885.70
4678756.37
13.18
3827375.65
13.20
3751501.90
16.77
1490867.62
6.91
I agree your transcriptomics indeed deepened your discussion. I have two technical questions:
- Have you compared with other de novo assemblers than rnaSpades or CD-HIT-EST? Any assembling performance data? Did you pre-process raw RNA sequences?
Coauthors Adrian Marchetti and Emily Speciale are part of an OCB Meta-Eukmoics working group that is doing an intercomparison of metatranscriptomic methods for microbial communities. Based on the work we've done thus far, there are no clear trends between the performance of different de novo assemblers (whether that be rnaSpades, Megahit, Trinity, etc.). Thus, we chose to use rnaSPAdes and CD-HIT-EST due to our familiarity with the softwares and their efficiency. We did pre-process raw reads using TrimGalore and FastQC.
(2) You extracted diatom reads/contigs for further downstream analysis. However, transcriptomic data might be unbiased and might have various reads/contigs from various organisms. What proportion did diatom data contribute to the total sequences? How much RNA data did you throw when only focusing on diatoms?
See the table above for information regarding the proportion of transcriptomic data.
We chose to focus on diatoms because we believe they dominant functional group among the photoautotrophs in this ecosystem and there is rich literature offering sufficient references to interpret diatom transcriptomic data. There are definitely substantial non-diatom organisms in our metatranscriptomics data that map to non-autotrophic organisms. When trying to look at broad expression trends, trophic mode can be hard to filter for, so we filtered for Bacillariophyta (aka diatom) mapped reads to represent the dominant phytoplankton response to various environmental conditions.
L110: Have you validated your nitrate sensor with your autoanalyzer outputs?
The sensor was recently recalibrated by Seabird prior to the cruise. Immediately prior to the cruise and during the cruise itself we validated the sensor against standards of sodium nitrate dissolved in MQ. We have added a note of this to the methods section (L35-36).
L139: How did you dissimilate/separate QA and QB reoxidation from your FRRf data? You can separate QA and QB if using an algal isolate, but was it appropriately separated in natural communities even with various Fe availabilities?
The FRRF method we applied is a single-turnover method which relies on the assumption that our full sequence of sub-saturating excitation pulses and following relaxation sequence leads to a single reduction and oxidation event of the collective (mixed assemblage) Qa pool. This assumption is based on the timing of the excitation sequence (127 sub-saturating pulses delivered over 250 us) which outpaces the reoxidation of Qb= -> PQH2 -> Cyt b6f which occurs over 4-5 ms and represents one of the rate limiting steps of photosynthetic light reactions. Multiple turnover methods with longer excitation time-scales (200 – 10,000 ms; e.g. PAM fluorometers) do record fluorescence signals associated with Qb reoxidation (see figure below).
Image credit: Chelsea Instruments. The multiple turnover shows a temporary pause in the rise of fluorescence when Qa is saturated. As Qa is reoxidized, fluorescence continues to increase until the entire electron transport chain is saturated. This is part of the reason Fv/Fm measured by PAM fluorometry is higher than FRRF.
L203: I understand NPQ fitting is quite tricky. Serôio and Lavaud (2011) thus discussed model performance on NPQ-E fittings. How were your single-component exponential fittings? Any stats outputs?
Stats added (L253): ‘Out of 91 curve fits, 95 % had R2 > 0.90 and 87% had R2 > 0.95.
A sample NPQ-PAR curve is included below.
L206: Why did you choose 46% light depth? You surely collected the same light depths at each station? Or simply light intensities at ~50 m almost similar at each station?
A primary objective of this cruise campaign was to collect measurements during a large scale Fe enrichment (+Fe) and depletion (+DFB) incubations. Light depths were sampled to mimic the on-deck incubators used for the Fe experiments. The Fe experiment is not part of this study, but influenced sampling strategies throughout the cruise for data comparison and compatibility. The bottom of the euphotic zone ranged between samples between 40-46m. Light depths were determined using the CTD-mounted PAR meter during our daily productivity casts enabling consistent sampling at 46% and 1% light levels.
L250: Any other source(s) of N2O other than photoinhibition?
Good Q, added some text to the methods and results section regarding uncertainty in NCP due to potential mixed layer nitrification, which was small.
L331-341: We note that several recent studies have observed nitrification within the euphotic zone, challenging the assumption that N2O production is limited to subsurface waters (Grundle, Juniper and Giesbrecht, 2013; Smith et al., 2014), and potentially leading to overestimates in our vertical mixing-corrected NCP estimates. Previous observations in the CCS reported a range of depth-integrated mixed layer nitrification rates between 0.3 – 2 mmol NH4+ m-2 d-1, resulting in consumption of 0.6 – 4 mmol O2 m-2 d-1 (Stephens et al., 2020). Following the approach of Izett et al. (2018) we used a range of realistic N2O:O2 stoichiometries to estimate potential upper and lower bounds of mixed layer N2O production. We determined mixed layer N2O production likely ranged between 0.09 – 0.23 mol N2O m-2 d-1, which would yield offsets in our final NCP estimates between 1.2 and 3.3 mmol O2 m-2 d-1. Total uncertainty due to sources of error in other derived parameters was determined by following Izett (2021).
L594-596: Uncertainty in vertical-mixing corrected NCP due potential mixed layer nitrification (see sect. 2.6) represented between 1.5 – 4.2% of our mean corrected NCP value.
L321 and elsewhere: How did you measure salinity? If from CTD, psu is no longer used. Just unitless.
Salinity was measured with a CTD and TSG. Psu units removed.
L354: In general, the Redfield ratio does not apply to surface waters I reckon because Redfield ratio 16 was measured in deep waters. I would rather recommend C: N allocation ratios in algal cells as discussed above.
Agreed that incorporating measurements of C:N in biomass (and Si) would have strengthened the manuscript. Unfortunately, that is not data we have on hand, so our discussion is limited to our measurements of dissolved N, P, and Si.
Figure 7: As you discuss, Cape Bianco had lower temperatures due to stronger upwelling. It is quite interesting but did this temperature difference affect (1) the enzymatic activity of xanthophyll pigment synthesis and consequently NPQ dynamics/xanthophyll cycle (XC)? and (2) transcriptomic regulation such as low-temperature adaptation, which overlayed Fe-regulated gene expression or TPM counts?
Regarding (1), NPQ correlated with SST in Table 1, which might be due to temperature-dependence XC or NPQ? Also, differences in the community structure might influence NPQ or other FRRf parameters?
Yes, NPQ, like Fv/Fm and other photophysiological parameters should vary with temperature and community structure. However, direct temperature effects on NPQ are expected to increase NPQ. Previous lab studies by Xu et al. (2012) (https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=phytoplankton+NPQ+cold+temperature&oq=phytoplankton+NPQ+cold+temper#d=gs_qabs&t=1743778443485&u=%23p%3DMa5utUAsm8gJ) have shown that NPQ increases under cold temperatures while While Yan et al. (2019) (https://onlinelibrary.wiley.com/doi/abs/10.1111/jpy.12846) demonstrated that Arctic phytoplankton enhanced NPQ photoprotective strategies under colder temperatures. Presumably, slower enzyme activity in the cold reduces capacity for protein repair and downstream electron transport, increasing susceptibility to photodamage or inhibition (Fanesi et al,.2016) (https://onlinelibrary.wiley.com/doi/abs/10.1111/fwb.12777). However, our data shows a positive correlation between NPQ and SST, opposite of the expected trend due to direct temperature effects. Here, SST acts as an indicator for nutrient-rich upwelling water. While the lower temperatures almost surely have an effect on enzymatic rates, it appears indirect sst effects, and therefore greater nutrient availability, are outweighing direct effects.
Table 1: In the 3rd row, Pmax should be ETRmax??
The Pmax refers to Pmax from photosynthesis-irradiance curves. Pmax is defined by eqn 3 in section 2.3 and the Pmax plotted in Fig 4b. Admittedly, Pmax/Vmax/Cmax terminology all got a bit muddled in section 4.3.1 and is likely the source of confusion here. All of the terminology throughout has been revised for consistency. Pmax is the maximum rate for ETR. Pmax-C is the maximum rate of carbon fixation.
Figure 5(c): GP at Cape Bianco is highly variable with numerous spiky up-downs compared with the peaceful Cape Mendocino. Any possible explanations for the difference? Or simply, large errors for estimation of GP in Cape Bianco?
Spiky data around May 31-June 1, indeed! Yet ETRPSII, which GP derives from is not spiky around those days. This indicates the variability is coming from the conversion factor from ETR to GP. To convert ETR to GP, we applied the following unit conversion:
Within the conversion factor above, the only variables that are not constant are the mixed layer depth and the mixed layer chlorophyll concentration. The chlorophyll data around Cape Blanco from May 31-June 1 was highly variable (see below) and that variability propagated into our GP estimates. Why was Chl data spikier around Cape Blanco compared to Cape Mendocino? Not sure exactly, but we were transiting across hydrographic fronts between offshore waters and the upwelling plume where a strong bloom was beginning to form.
L451: I wanna know x and y! Please add the exact numbers.
Apologies! Text updated: L568-569: “On average, the bottom depth of the euphotic zone was 14 12 meters deeper than the bottom depth of the mixed layer”
Figure 6: Quite interesting! I’m curious Stations 2 and 9 showed almost the same conversion factors between surface and subsurface. Any possible explanations? Because of intense vertical mixing at both stations?
It is a bit of a befuddling result – Strong similarities between the subsurface and surface samples would make sense in the context of strong upwelling. However, Station 2 (Cape Blanco) and 9 (Cape Mendocino) were very different from one another. Upwelling, estimated from the NOAA CUTI index, was not the same between the two capes (figure below).
Image caption: Study area colored by the NOAA CUTI upwelling index (1km^2 resolution) on May 31 (day of station 2 sampling) and June 10 (day of station 9 sampling). Star shape indicates our position on the day pictured. Indicating Station 2 sampling coincided with strong upwelling while station 9 sampling coincided with strong downwelling. Speculatively, strong vertical mixing can still explain the strong coherence between subsurface and surface samples at these stations although the prevailing direction of vertical transport differed.
Subsections 4.2 and 4.3 are a bit redundant for me even as an aquatic photosynthesis researcher although your discussion is quite interesting! It would be great if you could shorten these paragraphs.
We have re-edited these sections for brevity. While we hope the writing is more concise, we still kept lots of the fundamental information discussed in these sections since they provide critical context for our results.
Subsection 4.3.2: There is no mention of Fe availability. No need here? It would be more interesting if you could include just a few discussions on Fe availability as your manuscript title suggests.
We have updated L993-995: In contrast to , there was no significant differences in NCP:ETRPSII between Cape Blanco, Cape Mendocino, or offshore, suggesting limited effects of nutrient limitation on decoupling between ETRPSII and NCP.
We also added some text that explains our result, per Reviewer 1’s suggestion:
L1035-1047: Despite the inherent dependency of net oxygen production on gross oxygen production, the strength of the correlation between NCP and ETRPSII and the consistency of ETRPSII:NCPacross offshore, Cape Blanco and Cape Mendocino subregions is surprising given the vast suite of potential methodological and physiological sources of uncoupling (Fig 7). However, the derivations of NCP and GP both have similar dependencies on mixed layer Chl concentration. To obtain FRRF-derived GP estimates in comparable units of mmol O2 m-2 d-1, we multiplied in-situ ETRPSII by mixed layer Chl concentrations (Eq 5). While mixed layer Chl concentrations are not explicitly included in NCP calculations (Sect 2.6), biomass is expected to be a primary driver of bulk productivity. If Chl-normalized NCP is instead compared against GP expressed in units of mmol O2 Chl-1 d-1, the correlation between 24h binned and instantaneous NCP and ETRPSII estimates decrease to = 0.22 and 0.35, respectively. We therefore conclude that it remains challenging to derive gross and net carbon fluxes from FRRF measurements alone, but paired ETRPSII and Chl measurements can provide useful constraints for NCP estimates.
L685: OK, I would partially agree with your rough estimation using 400-700 chls/RCII. Is it also applicable to Fe-limited waters such as Cape Mednonico? If you look through other chls/RCII ratios in Fe-limited waters (e.g., Strzepek et al. 2019 with appendices and references therein).
Chl:RCII is expected to increase in Fe limited waters. Previously, Schuback et al. (2015) estimated Chl:RCII as 500 for Fe replete water and 700 for Fe limited water based on literature values from Greene et al., 1992.
We have added significant discussion surrounding Chl:RCII (L889-917) that will be of interest:
“In addition to non-linear electron transport, is also directly affected by the number of Chl energetically coupled to RCII (. Directly measuring requires specialized O2 flash yield instrumentation (Suggett et al., 2009), which was unavailable for this study. However, in-situ Chl concentrations, normalized to FRRF-derived proxies for [RCII] (∝ Fo/) following the approach of Oxborough et al. (2012), can be used to examine variability in between samples. With a known instrument calibration factor, Ka, either provided by instrument manufacturers or determined independently by O2 flash yield measurements, this approximation could be used to estimate the absolute value of .
It is well established that Chl:RCII () ratios increase under low light, to maximize light absorption (Greenbaum and Mauzerall, 1991). In our measurements, the proxy for varied significantly between sample depths, with higher at the bottom of the euphotic zone compared to surface depths, confirmed by a t-test comparison of population means (p << 0.01). Iron limitation is also expected to increase Chl:RCII. Although iron limitation lowers total cellular Chl content, Chl is more likely to be energetically coupled to RCII rather than PSI reaction centers (Greene et al., 1992). Accordingly, displayed a negative correlation with Fea1 expression in surface samples ( = -0.72, p <0.05, n = 9), which we used as a proxy for iron limitation. We thus conclude that Fe-stress likely contributed to variability in in addition to influencing non-linear electron transport. The hyperbolic relationship between carbon fixation and electron transport was unaffected by , which was assumed to be constant for individual samples throughout the course of photosynthesis-irradiance experiments (Appendix 4). However, this assumption may be violated under high light, due to photoinactivation of RCII (Campbell and Serôdio, 2020). A robust understanding of variability requires direct [RCII] measurements collected in parallel with ETRPSII and carbon fixation measurements.”
L731: Energy imbalance due to photosystem impairment can be indeed inferred from NPQNSV as you discuss. I would also suggest looking at qP or qL as described above.
Added in details on qP (referred to in our manuscript as F’q/F’m) per your suggestion (Table 1).
Citation: https://doi.org/10.5194/egusphere-2024-3812-AC2
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 909 | 170 | 27 | 1,106 | 27 | 38 |
- HTML: 909
- PDF: 170
- XML: 27
- Total: 1,106
- BibTeX: 27
- EndNote: 38
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Kate Schuler
Emily Speciale
Adrian Marchetti
Claire Till
Ralph Till
Philippe Tortell
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4712 KB) - Metadata XML
This paper investigates the inter-comparability of 3 different “currency” approaches to measuring primary productivity in the iron-limitation mosaic of the California Current System. In particular, it investigates the influence of iron availability/limitation on the inter-comparability of the approaches. This is useful in that it elucidates adaptations at the cellular scale that allow phytoplankton to cope with iron limitation, and it also shines a light on scenarios that do/do not allow a straightforward conversion from one approach to measuring primary productivity to another. The work is timely, interesting, well-structured and thought through, and the paper is well written, with care clearly taken to allow the reader to follow the arguments and methodologies. The figures are overall good and appropriate, and the main finding, i.e. that a fluorescence-based approach to measuring gross primary productivity is well correlated to daily integrated net community production is indeed surprising and interesting. Below are some very minor comments that hopefully will improve the presentation further.
L90: in the list of respiration pathways for NCP, shouldn’t bacterial respiration as well as respiration by larger animals be added to this list? See also Figure 7.
Figure 1: panels a and b are reversed in the caption.
L151: was PAR measured or assumed? If it was measured, how was it measured?
Eq 4: what is meant by “in situ PAR”? PAR averaged for the MLD? I’m struggling to make sense of the equation.
Eq 7: the superscript “B” – is it a designation (bio?) or an exponent? Can you make it a bit clearer by spelling it out?
L262: inconsistency regarding letter k in the fraction: it’s capitals here, but small-caps above (L253).
L266: since it’s the liquid fraction you’re interested in, should it be “filtered through” rather than “onto”?
Figure 2: could you indicate here what areas were used for the boxplots in Figure 3?
L365-368: These statements could distinguish a bit more between the capes – not all statements are true for both capes (Figure 4).
L378/79: this statement is not true for alpha at Cape Blanco – this whole section could be either a bit more refined, or throw in a few more “overall”s to indicate that what is said is not always true but a general trend.
L380: is beta shown anywhere?
L387: “Elevated values…” – this is only true for Cape Blanco. You said previously that both capes are upwelling, but to different degrees; maybe start early with a clear distinction between the two capes, but “upwelling” is not one of them, that’s true for both (see also L389).
L416/17: Is this the “in situ PAR” that confused me above? If so, make sure to call it that in the Figure.
L425-429: refer to Figure 5 here?
L440-444: maybe add a word here about the uncertainties (due to nitrification in the surface layer) that were mentioned in the methods? What order of magnitude are they?
L451: “x+/-y” wants to be filled with numbers
Figure 7 and its caption: Do you mean “spatial and time scales”? Only the time scales are mentioned in the Figure. Also, I can’t see pathway 7 anywhere in the Figure
L633: Did you mean to refer to Figure 6 here (rather than Figure 8)?
L653: “… with low Cmax” – this should be “…with low Pmax-C”, no?
Figure 8 and its caption: maybe make it clearer in the caption that blue and red also refer to different Fe-limitation regimes, or make a second panel where the dots are coloured by some measure of Fe limitation (Fv/Fm, sigma,… something like that)? This is a useful distinction, no? Also, Vmax (mentioned in the caption) is not shown, nor do I know what it would be based on what has been discussed?
L664: mention of Vmax again – should this be Ksat? It’s getting a bit confusing in this paragraph… if there is both a Vmax and a Ksat, then maybe a table of all the variables and their meaning would be useful?
L696/70: I would have thought that Cyt b559a would correlate with NPQ, and therefore expected a negative correlation with Pmax-C. Does this deserve a bit more discussion?
L678: Vmax again… and also, a question that comes up reading this paragraph and the preceding ones: How do you decide whether to discuss Vmax (Ksat) rather than Pmax-C? Is there a scheme to this? They seem somewhat interchangeable to me at this point. When you don’t mention one or the other, does it mean the correlation would have been insignificant? Again, maybe a word or two on this would be useful.
L678-80: This sentence is either missing a final section, or the “and” after the comma needs to go; also, please specify whether this increased coupling of Chl to RCII is specific to iron limitation or generally true.
L681: “increases in”? or “increased”? This whole paragraph could be a bit clearer - what does the literature say, what was measured here, what needs to be assumed, how does it all come together?
L704: Can you insert the actual ranges that have been observed by others?
Paragraph starting L698: The observed relationship between FRRf-derived GPP and NCP is a significant and surprising result (maybe you disagree, but then please lay out more clearly why it’s not surprising). Maybe a bit more discussion on why this relationship with NCP works here would not be amiss. Where/when do we expect this to work elsewhere (and where/when may it fall over)? What needs to be investigated next? Does it have anything to do with iron at all?
Figure 9 and caption: Does the correlation for panel b (rho = 0,92,L715) include the negative data point? How should one think about the negative data point, can it be explained, could it have to do with upwelling/downwelling? And what about the intercept for the line of best fit (-0.55), how to think about that? That it’s very close to zero and that’s good, or does it have some other meaning that’s worth spelling out? Also, the final sentence of the figure caption is a repeat, except for the r2. Regarding panel A of the figure: could it be improved by plotting the points as a heat map, to show where most of the data points fall?
L739: Regarding “fine-scale variability”: what scales are being resolved when we go to 24h binning?