Review for

Temperate marine macrophytes are highly variable sources of

Biogenic Volatile Organic Compounds - a comparative study

from the Baltic Sea and NE Atlantic

by Max Gräfnings1, Yuanyuan Luo2, Jian Zhao2, C 5 laudia L. Cara-Ortega3, Kirsten N. Fossum4, Frans Graeffe2, Lu Lei4, Dagmar B. Stengel3, Roseline C. Thakur2, Jurgita Ovadnevaite4, Mikael Ehn2, Camilla Gustafsson1

This study shows that temperate marine macrophytes emit a highly diverse and variable range of biogenic volatile organic compounds (BVOCs), with significant differences both between species and across environmental regions. It finds that these organisms—especially seagrass—can be substantial and previously underappreciated contributors to marine BVOC emissions, highlighting the need for more comprehensive data to better include them in global climate and atmospheric models.

Major strengths are:

The study identifies 166 BVOCs, far exceeding most previous studies that focused only on selected compounds or groups, providing a much broader picture of macrophyte emissions.

Unlike many earlier studies using destructive sampling or air-exposed measurements, this work quantifies underwater emission rates, making the results more representative of natural conditions.

The use of real-time, high-resolution mass spectrometry enables detection of very low emission rates and dynamic changes, improving accuracy compared to traditional GC-MS approaches.

By studying the same species in two contrasting environments (Baltic Sea vs. NE Atlantic), the study captures intraspecific variability, which is rarely addressed in previous work.

The study simultaneously analyzes compound identity, emission magnitude, and variability across species and environments, offering a more complete framework for understanding and modeling marine BVOC emissions.

Points to be addressed:

Measurements were conducted outside natural environments, which may alter organism behavior and BVOC emissions due to stress or handling, limiting how well results reflect real-world conditions.

The PTR-TOF method cannot distinguish isomers or fully confirm molecular identities, meaning many detected compounds are only tentatively assigned rather than definitively identified.

No single analytical method captures all compound types (e.g., some halocarbons are missed), so the reported emission profiles are not fully comprehensive.

Measurements were short-term and do not account for diel, seasonal, or long-term variability in emissions, which are known to influence BVOC dynamics.

High variability across species and regions, combined with relatively sparse sampling, makes it difficult to reliably upscale results to global BVOC budgets.

This study presents the BVOC composition and emission rates of three marine macrophytes (Fucus vesiculosus, Ulva intestinalis, and Zostera marina) across two regions (Ireland and Finland) and in two seasons (spring and summer). This manuscript is overall well written and the discussion is solid. I suggest providing greater methodological detail (see specific comments below). Otherwise, I believe this manuscript provides valuable information on BVOC composition and emission rates from macrophytes, whose cover has declined over the past century. There is therefore a growing need to accumulate data from similar studies for inclusion in marine BVOC budgets.

The article entitled "Temperate marine macrophytes are highly variable sources of BVOC - a comparative study from the Bastic Sea and NE Atlantic" evaluated BVOC emissions from 3 macrophytes (Zostera marina, Fucus vesiculosus and Ulva intestinalis) from two different locations (Ireland and Finland) by on line measurements through the use of Vocus PTR-TOF. The results presented in the paper are really interesting, highly qualitative and filled the gap of knowledge on marine BVOC. My main issue is about the statistical analyses (see below). 

- line 219 => it is not clear to me if Zostera marina were collected with their roots or not?

- line 227-228 => not sure that only a leaf rinsing is enough to remove all the epibiota from individuals. I would be more careful with this statement (providing maybe pictures of before/after the rinsing)

- line 268 => LPM is a bit weird to me. I would rather write L.min-1. 

- line 320 => I would say deposition or uptakes rather than consumption, just to be more careful. 

- For the PCA graphs, you could mix the A and B together with a selection of the main variable (by adding a threshold in your code) affecting the individual repartiton. Since there are a lot of information in this PCA, maybe in this new combined graph, no need to put the compounds where no elemental formula were found (but keep it for the statistical analyse of course). Also, I would rather put the real formula and not the protonated one. Finally, PCA is dedicated to parametric data which is often not the case for BVOC emissions (due to high variability), so most of the time, it is better to run a PLS-DA, more suitable for these types of data unless you have checked the normality and homoscedasticity of all compounds detected for all treatments.

- lines 380-393 => I don't really understand how you ran the PERMANOVA. Even though, it is non-parametric analyses, it works like MANOVA that handles several parametric variables. So according to your experimental design, you should run a two-ways PERMANOVA with species and régions as factors. Then you would obtain p-values for the effect of species, for the effect of region and the interaction between both factors. Then to highlight where are the différences (if you got significant effect of at least on factor), you should run a pairwise test. However, in the description of the results, it seems that PERMANOVA was used as a pairwise test which is not correct but maybe I misunderstood something. And you could add the results of the PERMANOVA in this PCA graph

- Also to illustrate your results in a very simple way, you could build Venn diagramm that quickly highlights shared and and unique compounds among treatments

- for the ANOVA on total BVOC emissions, I have the same issue than for the PERMANOVA. According to your experimental design, you should run a two-ways ANOVA with species and régions as factors followed by a posthoc test. 

- usually, emissions unity is written with . instead of / like ng.g-1.h-1. That needs to be changed throughout the manuscript

- throughout the manuscript, I would rather write the real molecular formula than the protonated one.

- no physiological data was measured during this experiment that coud better explain the differences obtain according to region? 

- lines 486 - 495 => here, I don't think that the use of GC-MS instead of PTR-MS is the only reason that you detect more compounds in your study. We could add at least two other explanation : 1) the method used were not the same. You used a dynamic sampling method by adding an airflow into the system while it was not the case at least in Coquin et al. 2024 and Bravos linares et al. 2010 (headspace + SPME) ; 2) you might have detected the emissions from epibiota (I don't think that only rinsing remove all the organisms from the leaves since some of them can be really well attached to it, see my comment above) while in Coquin et al. 2024 it was removed with a scalpel (not enough though to remove everyting I think). This part need to be developed.