The manuscript « Simulated terrestrial runoff shifts the metabolic balance of a coastal Mediterranean plankton community toward heterotrophy” evaluates the consequences of terrestrial runoff on planktonic communities in the coastal Mediterranean Thau Lagoon. For that the authors conducted in situ mesocosms experiment simulating terrestrial runoff by adding soil and measuring several chemical and biological parameters during almost 3 weeks.

The paper is easy to read, interesting and, I believe, deserve to be published in Biogeosciences.

Please find following some minor comments.

M&Ms

L76: Experiments were performed during 18 days, why?

L87: Turn-over rate of 3.5d-1. Does that mean that every 3.5 days water from the surface goes down to the bottom of the mesocosm and goes then up to the surface?

L92: The data that you presented here are the average of 2 mesocosms for the control and 2 mesocosms for the treatment. It is quite critical to use these data with so few replicates…

L101: “The mixture was left to mature for two weeks”. Any refs? Why 2 weeks?

1. 119: “Winkler Chl-a” , I think this is a typo error.

L121: You said that you left at least during 6 h the DO bottle of fixation. This is quite a lot. Normally, we should leave the DO bottles, once fixed with R1 and R2, to settle the precipitate until reaching 1/3 of the bottle volume. If, for any logistical reason, we have to analyze latter the fixed DO bottles, It is advised to be kept in the dark and under water. Was that the case here?

L135: Nutrient samples were stored at -20°C until analyses. However it is not recommended to keep samples for silicate analyses at -20°C but rather at 4°C. Indeed, freezing can cause polymerization that can biased the analyses and specifically for samples from coastal, estuarine waters and with lower salinity (cf. Aminot and Kerouel). Considering that, I think the silicate data has to be analyzed with great precaution.

How do you think, by the end of the experiment, the diminution of mesocosms volume can affect planktonic communities and thus your analyses?

L175-185: What is the % of error of this method? Its reproducibility and inter-repeatability?

Results

Figure 1: How do you explain the great drop of DLI at day 12 and 13?

Figure 2: BA had opposite trends between control and treatment. How do you explain it? Why did you not analyze the phytoplanktonic abundance?

Figure4: Did you check CR/Chl or CR/BA?

Discussion

L382-387: Considering the mixing of the mesocosms waters (turn-over of 3.5 days), do you think that sedimentation is important?

L390: There is no relationship between CR and BA, so we cannot think that planktonic bacterial respiration represent a major part of community respiration here.

L407: Do you think that Mediterranean coastal lagoons are really net O2 producers today? Any refs?

Supplement information

Rnight = (mean of NCP during night period)*duration of night period*60 (Soulié et al. 2021.

The work by Soulié et al. “Simulated terrestrial runoff shifts the metabolic balance of a coastal Mediterranean plankton community toward heterotrophy” investigated the consequences of terrestrial runoff on plankton communities and some biological processes in the Thau lagoon on the Mediterranean coast. The results come from an in situ mesocosm experiment in which terrestrial runoff was simulated by adding soil and various chemical and biological variables were analyzed for 18 days.

The paper is well written, focused and interesting and should be published in Biogeosciences after revision.

A major criticism of this work is the fact that no real phytoplankton data are presented. All results were derived from Chl-a sensor data, including phytoplankton growth and loss rates. I know this is becoming more common and accepted lately, but similarly to Chl-a fluorescence and oxygen sensors data, for which some calibrations were done, actual phytoplankton counts and identifications should have been done on at least some samples to check if there is a match between Chl-a and microphytoplankton. Since flow cytometry was used in the paper to estimate heterotrophic bacterial abundance, the same method could have been used to assess the smaller phototrophic picoplankton. It is a pity that there is no information on the actual composition of the phytoplankton, as this is the topic of the article. If the authors have these (already published or unpublished) results, I think it would be a great addition to this publication to at least mention them in the discussion.

Introduction

I think it would be easier for the reader if some scientific questions or hypotheses were added at the end of the introduction section to structure the discussion.

M&Ms

Experimental design

It is not stated what the final volume in each mesocosm is. It is also not clear if the mesocosms are sealed at the top and what the bottom is like. Are the mesocosms open or sealed at the bottom? Are they floating? Please add this information.

L76: The experiments were conducted for 18 days. In my opinion, this is a very long time to test the effects of a flash flood on the coastal area. There is a strong mixing effect by the seawater, and even in a confined area like a lagoon, the terrestrial runoff will eventually be diluted in a few days. The choice of the duration of the experiment should be discussed.

L104: Why was the sampled soil left to mature for two weeks? Why was this step necessary? In a flash flood there is no such step, the soil is washed away directly by the heavy rain and transported into a river and eventually into an estuary. Flash floods do not last 14 days, this is unclear and should be better explained.

Discussion

What I miss here is a discussion of the ecological implications of these results at the lagoon mesoscale. In particular, a detailed discussion of the trophic state of the lagoon and a comparison with other Mediterranean lagoons.