PREreview of Chlorophyll fluorescence: How the quality of information about PAM instrument parameters may affect our research

The manuscript by Nies et al. demonstrates how changing pulse amplitude modulation (PAM) parameters can affect non-photochemical quenching (NPQ) and photosystem II yield (ՓPSII). Using in silico simulations of PAM experiments, the authors illustrate how NPQ and ՓPSII are affected by varying: i) the delay between measurement of maximal fluorescence Fm and the onset of the actinic light (or between turning off AL and measurement of Fm’’ in the dark), ii) the intensity of the actinic light, iii) the frequency of the saturating pulses, iv) and their duration. Nies et al. finish by validating their in silico model, and suggesting that scientists using PAM must provide the details of all of the parameters listed above in the methods section of any publications to allow their experiments to be accurately reproduced and modeled. We enjoyed this manuscript, however, we have some comments and suggestions for improvement, listed below.

Major comments

- We suggest moving part of the model validation section of the results, shown in Figures 8 (and 9), to the start of the manuscript. This rearrangement would show the reader that the mathematical model used in the in silico simulations can accurately reproduce experimental data, before the parameter-dependent changes to NPQ and ՓPSII are simulated. In the current arrangement, the reader needs to have prior knowledge that the changes in NPQ and ՓPSII shown in the simulations are accurate, before the herein updated model has been validated.

- Fig8. Regarding the validation of the mathematical model by comparing to experimental PAM measurements with different SP durations, or different delays of AL onset from Fm measurement, with the simulated data: what is the rationale for choosing these, how about testing the other parameters such as AL intensity and frequency of SP? Please comment on the impact of the different parameters on e.g. the NPQ measurement and rank them by stronger to lower effect based on your simulations and experiments. Also a historical perspective/physiological relevance of delaying the SP from actinic onset would be welcome! How about giving recommendation to researchers in the field to have Fm determination/SP right at onset of illumination, with no delay, to prevent further confusion (and similarly have the final SP in AL on, followed by AL off with no delay).

- Line 326. Regarding the use of another model of photosynthesis, we found this very interesting and suggest that a comparison of the simulations generated by the two mathematical models using the same set of parameters be included as a main or supplemental figure, and its description be included in the results section. The GitLab link (line 330) doesn’t specify which exact file to look at.

- Line 127. “We have used 500 µmol · s−1m−2 as the default light intensity of AL.” For simulations, an intensity of 500 µmol m−2s−1 was used, but for experiments (line 152) “The intensity of red AL was set at approx. 457 µmol m−2s−1”. We understand that matching the actinic light during the experiment to 500 µmol m−2s−1 cannot be possible, alternatively we suggest that the simulations be carried out at 457 µmol m−2s−1 for sake of consistency. Importantly, is 457 µmol m−2 s−1 the value given by the manufacturer for the chosen setting, and did you measure it to confirm its value? (depending on instrument calibration, usage and age, the light output can be different than set)

- Line 204, 205. “The calculated steady-state NPQ values are higher for SP intensities below 3000 µmol·s−1 m−2”, according to Fig.5, it seems that the threshold is rather 2000, than 3000 (or 4000).

- Fig7. To test the “actinic effect” of SP duration, we would suggest to perform a simulation with AL=100 µmol m−2 s−1 AL and/or AL=0 to check whether SP themselves can induce NPQ.According to Fig8A (experimental), it seems that at 0.8s, NPQ is indeed slightly higher than with shorter SP duration.

- Line 370, a necessary addition would be to list here, or write a template of, what you suggest for minimum information is needed as standard for the community. It could be similar to Table 2, and needs to include duration of AL on, off and AL quality.

Minor comments

- Line 46. “Allow”, should be “allows”

- Line 75. “Groups but also” should be “groups experimentally, but also”

- Line 115. Replace higher by vascular.

- Line 140. 26C is higher than standard temperature for Arabidopsis growth (22C), what’s the rationale for choosing this temperature?

- Line 150. Define Fv and explain if the 5s of far red light is turned on at the very beginning of the experiment i.e. before time 0.

- Line 153. “default settings (10)”, specify “set at value of” 10. We suggest writing a small table with these parameters (see major comment).

- Line 161. Which leaf did you choose, younger or older? This information is important to state, see differences with leaf age for example in Bielczynski et al. Plant Phys (2017) doi: 10.1104/pp.17.00904.

- Line 173-174. We suggest that the SP time points are moved to the methods section.

- Line 185-186. “In the upper panel….derived NPQ and ՓPSII”, this whole sentence can be removed as it should be clear from the figure legend.

- Line 211. “Far more” how many did you look at?

- Fig. 6. “6A and 6A”. Should be “6A and 6B”

- Line 234. “Switching on AL with the first SP in light-triggered after 1 s” suggest rewording as it was unclear what light-triggered means.

- Line 241. The observed effect is likely due to the total conversion of zeaxanthin to violaxanthin for long periods of dark-adaptation.

- Line 243. Suggest changing “whereas” should be “however” as it is clearer.

- Line 256. Define PMST.

- Line 264-268. We suggest moving this block of text to the discussion section.

- Line 264. “AL is another important information” should be “AL is another important piece of information”

- Fig. 8B and 8D. As the simulated curves seem to all overlap, and often in this study we look for fine nuances between data, we think it would be beneficial to read the simulated curve superimposed on top of the experimental data allowing a fair comparison and analysis between the two types of data. Displaying the same graphs at a larger scale would help to read them.

To help in this, we propose Figure 8 to be divided in two figures, since Fig. 8A-D is related to “SP experiment” while Fig. 8E-H is related to the “delay experiment”. This would allow the size of the panels to be increased to help the reader interpret the data.

- Fig. 8F and 8H. Plot titles “Delay NPQ/ՓPSII Sim lation”, should be “Delay NPQ/ՓPSII Simulation”

- Fig. 9 seems to be redundant as the reader should be able to observe the difference between the two independent experiments by comparing Figure 8A and 8E. We therefore suggest that Fig. 9 be removed.

- Line 286. “Measurements are” should be “measurements have been”

- Line 289-303. We suggest moving this block of text to the introduction section

- Line 324. Replace “many” by “all”!

- Line 351. “Agreements” should be “agreement”

- Line 361-372. We feel that the points made in this block of text have already been made earlier in the manuscript and repeated several times. Therefore this block of text can probably be omitted as it is redundant.

- General comments concerning the figures: we suggest adding dark/light bars to the top of most plots in Figures 3B-C, 4B-C, 6A-D, 7A-B, 8A-H; as it would improve the readability/interpretation of the plotted data. Fig. 2-8, figure identifier letters are presented in a different font style than the rest of the text, throughout the document. While we recognize them to be hyperlinks, we think font style should be uniform.

Jack Forsman and Andre Graca (Umeå University) - not prompted by a journal; this review was written within a preprint journal club with input from group discussion including Alizée Malnoë, Jingfang Hao, Maria Paola Puggioni, Pierrick Bru, Aurélie Crepin, Wolfgang Schröder.