the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Apr 2026
Impact of light, turbulent mixing, and nutrients on phytoplankton photoinhibition in the euphotic zone from the subtropical to subpolar North Atlantic
Abstract. Photoinhibition, a condition caused when high photon fluxes lead to the rate of electron transport from light receptors in photosystem II exceeding a cell’s photochemical capacity, causes damage to the photosynthetic apparatus and reduces photosynthesis. The balance between damage and repair processes in cells determines the degree of photoinhibition (PI). PI is here shown to be ubiquitous in natural phytoplankton communities in the North Atlantic. PI effects were determined as the difference between the maximum photosynthetic quantum yield (Fv/Fm) recorded at the time of sampling and following a 4 h low-light incubation. Samples from the euphotic zone in the North Sea showed significant daily variation of PI extending to depths of 20 m. Principal component analysis identified a strong correlation between PI and light availability, whereas nutrient distributions and other environmental variables were less influential on PI. Furthermore, an almost immediate response of PI to changes in light intensity was demonstrated. A light-dependent linear regression model of PI for the northern North Sea explained a significant increase in PI (R2 = 0.4) with intensity of insolation. PI in samples from the euphotic zone in the subpolar Irminger Sea and the subtropical Sargasso Sea also showed significantly elevated PI at higher light levels. Our results show that PI in natural phytoplankton communities is closely related to, and occurs almost immediately, in response to changes in insolation. The sensitivity of phytoplankton communities’ photo-chemical capacity to insolation changes may help to explain the wide variation in photosynthetic parameters reported in nature and should be taken into account when estimating ocean primary production.
- Preprint
(2601 KB) - Metadata XML
-
Supplement
(53 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1718', Patrick Neale, 19 Jun 2026
-
AC1: 'Reply on RC1', Jørgen Bendtsen, 05 Aug 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-1718/egusphere-2026-1718-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Jørgen Bendtsen, 05 Aug 2026
-
RC2: 'Comment on egusphere-2026-1718', Anonymous Referee #2, 22 Sep 2026
This manuscript presents an interesting field dataset examining light-associated changes in phytoplankton photophysiology across contrasting North Atlantic environments. I particularly appreciated the extensive observations from the North Sea, including the time-series station, and the combination of physiological measurements with information on the physical environment. The relationship between the PI metric and irradiance, particularly in the upper water column, is clear and potentially useful.
My main comments concern the interpretation of this relationship rather than the observations themselves. In a few places, I think the conclusions go somewhat beyond what can be directly inferred from the measurements, particularly regarding the physiological meaning of the recovery in Fv/Fm, the short response time, and the relative importance of light compared with other environmental variables. I think these points can mostly be addressed by clarification and somewhat more cautious wording.
- Interpretation of the PI metric
The approach assumes that the increase in Fv/Fm during the 4-h low-light incubation represents recovery from photoinhibition, presumably through repair of PSII. This seems a reasonable operational approach based on From et al. (2014), but I think the manuscript sometimes treats this interpretation more directly than the measurements allow. Damage and repair rates are not measured independently here.
One observation that seems relevant is that the corrected Fv/Fm_c itself changes during the day. Since this value provides the recovered reference state used to calculate relative PI (Pi_PI), I think its diel variability deserves some discussion.
I therefore suggest describing Pi_PI more explicitly as an operational estimate of PI based on low-light recovery of Fv/Fm, and being somewhat cautious when interpreting it specifically in terms of PSII damage and repair.
- Response time to changes in light
The light-history analysis shows that PI is more closely related to recent irradiance than to irradiance averaged over several hours. I am less convinced, however, that the analysis resolves a physiological response time of approximately 20–30 min. Instantaneous and 10–30 min PAR give very similar R2 values, while averaging over longer periods will also smooth the diel light cycle.
There is also a clear depth dependence in Table 1. For samples from 0–10 m, R2 is approximately 0.44–0.45 for instantaneous through 30-min PAR, whereas below 10 m it is only about 0.03. Thus, the relationship for the complete dataset seems to be mainly driven by the upper water column.
I would therefore emphasize the result that PI is closely associated with recent light exposure in the upper layer, while being somewhat more cautious with the conclusion of a <30-min physiological response time, particularly where this is described as an “almost immediate” response.
- PCA and broader interpretation
The PCA nicely illustrates the environmental structure of the dataset and the close covariance between PAR and PI. At the same time, PC1 also seems to represent an important vertical gradient: PAR, PI and temperature covary toward the surface, while nutrients show the opposite pattern.
For this reason, I am not sure that the PCA alone demonstrates that light is the “dominating variable” controlling PI or that nutrients and other variables are less influential. I would probably describe this result more simply as a strong covariance between PAR and PI, unless the authors wish to separate the effects of the different environmental variables more explicitly.
I would also be somewhat cautious with some of the broader implications. The Irminger and Sargasso Sea observations are valuable because they show similar light-associated responses in contrasting environments, although there are also deviations from the North Sea relationship. In addition, the measurements concern maximum PSII photochemical yield rather than carbon fixation directly. The implications for primary production are certainly worth discussing, but statements concerning ubiquitous PI or constraints on photosynthesis across large ocean regions seem somewhat broader than can be demonstrated from the present observations.
Specific comments
- Incubation conditions: Please provide some additional information on the 4-h low-light incubations, particularly incubation temperature and whether it was controlled relative to in situ conditions. Was 4 h known to represent a stable recovery endpoint across the different regions/communities?
- Sample size and supplementary data: I may have misunderstood the structure of the supplementary datasets, but I could not reconcile the sample size reported for the North Sea PAR–PI regression. The manuscript reports n = 281 for observations with PAR > 50, whereas I find 120 observations satisfying this criterion in the supplied .dat file. Interestingly, using these observations gives essentially the reported regression coefficients and R2. Could the authors please clarify which observations were used for Eq. 3 and Table 1 and whether the complete PAR dataset used in these analyses is included in the Supplement?
- Statistical structure: Related to the previous point, were observations from different depths, repeated casts and the time-series station treated as statistically independent in the regression? Some clarification of the observational structure would be useful.
- Range of the PAR–PI relationship: Equation 3 was fitted only to observations with PAR > 50 micromol photons m-2 s-1. I suggest making this range clear when the relationship is subsequently discussed or applied, particularly since the positive intercept should probably not be interpreted at very low PAR.
- Regional comparison: Different fluorometry systems were used in the North Sea, Irminger Sea and Sargasso Sea. A brief comment on whether this affects quantitative comparison of Fv/Fm and Pi_PI among regions would be helpful.
- Terminology: In a few places, Delta Fv/Fm is described as a reduction in “photosynthetic activity”. Since carbon fixation was not measured directly, “reduction in Fv/Fm” or “reduction in maximum PSII photochemical yield” might be more precise.
- Minor corrections: Figure 4 appears to use Pi_NPQ, whereas I assume this should be Pi_PI. I also noticed a few minor inconsistencies in the reference list and DOI formatting that could be checked during revision.
Citation: https://doi.org/10.5194/egusphere-2026-1718-RC2 - AC2: 'Reply on RC2', Jørgen Bendtsen, 06 Oct 2026
-
RC3: 'Comment on egusphere-2026-1718', Anonymous Referee #3, 20 Oct 2026
This manuscript presents an interesting field study of light-associated variation in phytoplankton photophysiology in the North Sea, Irminger Sea, and subtropical Sargasso Sea, with the most extensive sampling conducted in the North Sea. The authors assess photoinhibition by comparing maximum PSII photochemical quantum yield before and after four hours of low-light incubation. The observed relationship between irradiance and the recovery-based photoinhibition metric, particularly in the upper water column, is clear and potentially useful for understanding variability in photosynthetic physiology and its implications for primary production estimates. A particular strength is the combination of paired physiological measurements, spatial sampling, diel observations, and measurements of the physical environment.
I offer the following comments to improve the clarity, reproducibility, and interpretation of the study.
- Notation: Please consider adding a table defining the principal symbols, their units where applicable, and their physiological interpretations. In particular, distinguishing the initial and corrected (F_v/F_m), their difference, and the normalized photoinhibition metric would help readers follow the analyses.
- Dataset provenance, documentation, and reproducibility (Sections 2.1–2.3 and data availability statement):
- Please clarify which observations were collected specifically for this study and which, if any, were previously reported or analysed elsewhere. For previously published observations, identifying the relevant publications and explaining their use here would help establish the provenance of the dataset.
- The supplementary spreadsheet provides station coordinates and fluorescence measurements, while the accompanying .dat file includes North Sea environmental variables such as nutrients, temperature, salinity, fluorescence, and PAR. A clearer description of how these files relate to one another would improve their accessibility. In particular, please identify the observations used in the PCA, Equation 3, and the light-history analyses, including any exclusions or missing-data treatment.
- I encourage the authors to provide a README or data dictionary explaining identifiers, units, missing-value codes, and links between files, together with the analysis code and any additional inputs needed to reproduce the results. Archiving these materials in a repository such as Zenodo and citing a persistent DOI would facilitate reproducibility and future reuse.
- Sampling depths and sample sizes (Section 2.1)
For the statement describing sampling “at 3 stations at multiple depths” in the Irminger Sea, please specify the sampled depth range. A table summarizing the number of stations, depth ranges, and numbers of paired fluorescence measurements for each region would also be helpful. Where feasible, please distinguish the number of stations, sampling events, and individual observations, particularly for the time-series stations. - Section 3.2 and Discussion:
- The model-derived reduction of approximately 45% in maximum PSII photochemical yield at a PAR of approximately 1000 µmol photons m⁻² s⁻¹ is an interesting result.
- The discussion could benefit from considering complementary evidence from carbon-based photosynthesis–irradiance experiments, such as Amirian et al. (2025; https://doi.org/10.1038/s43247-025-02686-3 & 2026; https://doi.org/10.5194/essd-2026-651), where relevant. These studies may provide useful context for discussing photoinhibition in natural phytoplankton communities. However, fluorescence-derived changes in maximum PSII quantum yield and high-light reductions in carbon fixation represent different quantities. A brief discussion of how these approaches complement one another would strengthen the manuscript.
- Interpretation of the PCA: I share Referee #2’s concern about using the PCA to infer the relative influence of light and nutrients.
- Quantitative summary of Figure 7
Please consider supplementing Figure 7 with the numbers and percentages of observations for which (\Delta F_v/F_m) is positive, zero, or negative, separately for the two depth groups shown. This could be presented in the caption or a supplementary table. Please also discuss the possible explanations for negative values, including measurement variability and the possibility that the incubation irradiance adversely affects samples acclimated to lower light. These percentages should be described as the frequency of positive or negative measured responses; a value above zero does not, by itself, establish statistically significant photoinhibition in an individual sample. - Figure 8
- Several aspects of this figure suggest opportunities to refine the statistical model. First, samples from different stations may respond differently to changes in PAR. A linear mixed-effects model with station-specific intercepts and, where supported by sufficient within-station observations, slopes could account for this variation and facilitate comparisons among stations.
- Second, have the authors explored transformations of PAR and the response variable? Although a conventional log–log transformation cannot accommodate zero or negative values, a log1p transformation, defined as (\ln(1+x)), can accommodate zeros and negative values greater than −1. Such a transformation may help address nonlinearity or unequal residual variance, although its suitability should be assessed using residual diagnostics rather than assumed to improve model performance.
Citation: https://doi.org/10.5194/egusphere-2026-1718-RC3 -
AC3: 'Reply on RC3', Jørgen Bendtsen, 02 Nov 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-1718/egusphere-2026-1718-AC3-supplement.pdf
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 543 | 159 | 70 | 772 | 68 | 50 | 59 |
- HTML: 543
- PDF: 159
- XML: 70
- Total: 772
- Supplement: 68
- BibTeX: 50
- EndNote: 59
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors have systematically determined the increment between dark-adapted Fv/Fm and Fv/Fm after a 4 hour incubation in low light in profiles of discrete samples in the mixed layer of several areas in the Atlantic Ocean. The increment is interpreted as the degree to which the function of PSII has been inhibited by in situ irradiance exposure. This relationship is shown in greatest detail in field measurements in the North Sea, from which a regression relationship is developed between in situ PAR and the incubation-based increase in Fv/Fm, either as an absolute change or relative to post-incubation Fv/Fm. The greatest changes are observed near the surface and around mid-day under cloud-free conditions. Smaller datasets show that similar phenomena can be observed in the sub-polar (Irminger Sea) and sub-tropical (Sargasso Sea) ocean, though the relationship with PAR is not the same as observed in the North Sea.
The exposure-sensitivity of Fv/Fm to high light and the recovery under low-light is a well known and widely observed characteristic of phytoplankton sampled from open waters, so the observations reported are not surprising. Nevertheless, their field observations constitute one of the most extensive datasets demonstrating its occurrence in a variety of ocean regimes and validates the suggestions of the earlier From et al (2014) study that these incubations provide not only a “corrected” Fv/Fm but also a widely applicable indicator of photoinhibition. An added feature of the dataset are the accompanying observations of microscale turbulence and standard oceanographic measurements.
A general comment is that the N. Sea field studies, like many, were made under what look like relatively calm conditions. Thus, the observations relating to light history probably do not consider the effect of large-scale vertical transport as could occur due, e.g. Langmuir circulation cells, on the profile of photoinhibition and how it would be affected by mixing.
The authors might want to consider changing the title, the current one doesn’t convey that the majority of the data is from the North Sea. Neither is enough said about nutrients to justify its inclusion in the title. An alternative might be: Phytoplankton photoinhibition and turbulent mixing in the euphotic zone of the North Sea and selected regions of the subtropical and subpolar Atlantic Ocean.
Minor Comments:
Section 2.2 turbulence: – Was this only measured in the North Sea?
Section 2.4 PAR: OK to assume that albedo is more-or-less the same over a given area, but albedo will vary through the day, this may have bearing on the regressions evaluating the effect of light history.
Section 2.5 Low-light incubation: In some species, including cyanobacteria and diatoms (e.g. Goss and Jakob 2010 ), dark adaptation does not completely reverse NPQ due to dark electron flow through the PQ pool. This can be reversed by giving the samples a short exposure to very low light (< 10 micromol m-2 s-1) just before the Fv/Fm determination. Did the authors check for this in their samples?
For the Sargasso Sea, From et al (2014) noted that there was a bottle effect which progressively reduced Fv/Fm with time. They made a bottle effect correction to their data, was this similarly applied in this study?
From et al (2014) also noted that incubation period needed for recovery varies between sites, a primary variable being temperature. What was the incubation temperature in the Irminger Sea samples and was it verified that recovery had reached an asymptote after 4 hours?
Line 146: Note typo 0^(-3) I assume it should be 10^(-3)
Supplemental Information – It would be good to include the full field data like that provided for Vermix also for the Sargasso Sea and Irminger Sea
Reference:
Goss, R., Jakob, T. Regulation and function of xanthophyll cycle-dependent photoprotection in algae. Photosynth Res 106, 103–122 (2010). https://doi.org/10.1007/s11120-010-9536-x