the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 28 Jul 2025
A first predictive mechanistic model of cold-water coral biomass and respiration based on physiology, hydrodynamics, and organic matter transport
Abstract. Cold-water corals form complex three-dimensional structures on the seafloor, providing habitat for numerous species and act as a carbon cycling hotspot in the deep-sea. The distribution of those important ecosystems is often predicted by statistical habitat suitability models, using variables such as terrain characteristics, temperature, salinity, and surface productivity. While useful, these models do not provide a mechanistic understanding of the processes that facilitate cold-water coral occurrence, and how this may change in the future. Here, we present the results of a mechanistic process-based model in which coral biomass and respiration are predicted from a 3D coupled transport-reaction-model for south-east Rockall Bank (NE Atlantic Ocean). Hydrodynamic forcing is provided by a high-resolution Regional Ocean Modelling System (ROMS) model, which drives the transport of reactive suspended particulate organic matter in the region. The physiological cold-water coral model, with coral food uptake, assimilation, and respiration as key variables and with model parameters estimated from available experimental report, is coupled to the reactive transport model of suspended particulate organic matter. Model predictions agree with coral reef biomass and respiration observations in the study area and coral occurrences comply with predictions from previously published habitat suitability models. Cold-water coral biomass was mainly predicted on coral mounds and ridges in the area. Filter feeding activity by cold-water corals proved to strongly deplete food particles in the bottom waters. Replenishment of food particles by tidal currents was therefore vital for cold-water coral growth. This mechanistic modelling approach has the advantage over statistical and machine learning-based predictions that it can be used to obtain an understanding of the effect of changing environmental conditions such as ocean temperature, surface production export, or ocean currents on cold-water coral biomass distribution and can be applied to other study areas and/or species.
- Preprint
(7354 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-3385', Anonymous Referee #1, 27 Sep 2025
-
RC2: 'Comment on egusphere-2025-3385', Wilder Greenman, 30 Oct 2025
General comments
The paper outlines the development of a new mechanistic model predicting coral biomass and respiration rate at Rockall Bank. Model data is compared to observational data and habitat suitability models from Rockall Bank to assess how the new model performs. As someone who is interested in food delivery to deep-water coral habitats I am very excited about the future possibilities of this type of work. I suggest minor revisions, which are outlines below. If this revisions can be addressed, I would happily support this for publication and look forward to seeing what mechanistic modelling of corals can tell us about nutrient cycling.
Specific comments
It would greatly benefit the strength of the paper if the authors could find a way to compare model results to observational data. Maybe an example could involve taking CWC biomass (mmol C/m2) as a ratio of carrying capacity to get a number that is comparable to percentage cover? A few lines of text going over this could replace the section where authors say results are challenging to compare. In doing so, they could elaborate on what type of substrate/conditions/depths the model is most useful for.
The model shows quite good agreement with a published habitat suitability model which is good to see, but more explanation on why the presented model should be chosen over a habitat suitability model moving forward could be explored. An example is given on line 585 where the effect of rising T could be indirectly included in the new model, but presumably T could also be directly updated in a habitat suitability model to predict the effect of warming T? It seems like exploring changes to food availability is a strength of this mechanistic model so expanding on this, or something similar would be very interesting.
Scientific questions
83 – what is meant by interactions?
85 – enhanced – This word choice makes it sound like corals increase the amount of OM. But I assume it’s meant to convey that OM concentrations are higher around the reef relative to nearby sediments because of their ability to retain nutrients? Maybe rephrase as “availability is higher on the reefs related to around them”? Or something like this?
166 – Why is suspended POM selected over sinking POM? Suspended POM is usually more refractory than sinking POC? Sinking POM would presumably have a faster settling rate and so could change model outputs. A line or two on the distinction and the decision for the model would be informative.
In the text, could you expand on any environmental factors that could explain why the model has difficulty matching VT6 in figure 8? Such a large presence of corals from 0-250 m coincides with very low biomass compared to what is predicted in VT2, VT5. Line 386 says the model agrees with observational data but they don’t match that well for some sections. What is it about VT2 that leads to such high values? Some specifics on the sites could be useful beyond mentioned model issues (depth of the model and patchy resolution).
Technical revisions
Line 9 – space between 2 and C which other affiliations don’t have.
Line 22 – The distribution of “these”
Line 25 –30 It’s a bit unclear how many different models are used when reading the abstract. Rephrasing this so it’s clearer would help the reader.
Line 30 – available experimental reports - plural.
Line 32 – occurrences “comply” – Odd word choice?
Line 72 – suggest revising sentence. As stated, this is a long sentence and it might be clearer/more accurate to take the latter half “and new tools and models are needed to understand how CWCs will be influenced by a changing marine environment” and rephrase it to say that new tools and models represent a way to offset or supplement, etc. the challenges of physically sampling.
Line 78 – provided with observation of CWC “habitats”? add word possibly? Or is it supposed to be about the physical traits of corals? As written, it’s a bit ambiguous.
107 – consider adding Girard et al., 2022 - https://doi.org/10.1098/rspb.2022.1033
141 – “biogenic soft sediment” < what does the descriptor “soft” add? As if soft rock? (i.e. sedimentary over metamorphic or igneous?). If so, I think this is covered by it being "biogenic". Or soft as in unconsolidated? Consider either removing or replacing word “soft”? If the latter is the intended description, then maybe “unconsolidated biogenic sediment”?
143 – It would be great if you explicitly state the relief of the key mounds above the biogenic sediments.
161 – possible suggested re-write: “The keystone species D. pertusa is our selected model species because it provides habitat/and contribute to is important to reef metabolism.
162 “used as (the?) model species” missing word?
193 – variably called organic matter transport model, organic matter reactive-transport model, or reactive transport model. Should be consistent throughout.
198 – “/” which should be a “.”?
201 – consider replacing word “representable” with “representative”.
207 – it would be nice if this formula were in a slightly smaller font and could be on one line.
210 – No POCC in formula – is the extra C a type?
230 – biomass “is” calculated? Typo?
253 – and consists “of” 1.36? typo?
254 – D. pertusum comprises of 2.12 <- should either be “comprises 2.12” or “consists of 2.12”.
294 – change in font size
394 – missing word? Due “to” a patchy distribution
420 – typo – where framework is “built” up
426/27 – font change for references
Figures
As a general comment for all map figures, it would be informative to have contour interval either on the map or in the caption. Contour lines are labeled in figure 10C but nowhere else I believe.
Figure 1
Missing scale bars, contour labels and geographic labels throughout Figure 1. Figure 1A is not useful if the reader is not familiar with the region. Ideally, UK and Ireland could be labeled. B and C – what are the intervals of the contour lines? Either add to figure or in caption. C) Latitude label is mostly covered and if there was meant to be a longitude label it is missing.
It's also a bit confusing that red arrows convey current trajectory and as a label pointing at the mound. Maybe change the colour of one? Or possibly remove the arrowhead from the Haas mound label?
Figure 2
First word should be capitalised.
Figure 3
Could you provide a scale quantifying the shade of green used? Presumably darker means higher concentration of POC?
Figure 4
I think coral region and coral presence lines are the same? Could you simplify this to one label so the figure and caption are consistent?
Figure 6
Consider replacing this with panel D from Figure 10. It would be more useful to show how the modeled data compares with the habitat suitability model since that is a focus of the paper.
Figure 7
Why does coral presence reach 40-50% in panel A, but corals are nearly absent in Figure 8F? Shouldn’t they both come from video 7?
Figure 8
Panel G – should these numbers be 1-7? Not all 1?
Citation: https://doi.org/10.5194/egusphere-2025-3385-RC2 -
RC3: 'Comment on egusphere-2025-3385', Maria Rakka, 03 Nov 2025
General comments
de Froe et al., present a very interesting and novel study using mechanistic modelling to predict coral biomass, and respiration. The work is highly interdisciplinary, effectively integrating physical oceanography, ecophysiology, and ecology. To my knowledge, this is one of the first mechanistic models applied to a deep-sea benthic species, and among the first to incorporate feedback loops between environmental and biological processes in deep-sea ecological modelling. The study is therefore both original and of high scientific quality. The materials and methods are clearly described and concise, and the authors demonstrate a strong understanding of coral feeding biology and ecophysiology, which are well integrated into the model. Model limitations are also well explained.
I suggest that the authors emphasize the innovative aspects of their work more strongly. A brief paragraph in the introduction about the use of mechanistic models in deep-sea benthic ecology would help set the context for non deep-sea specialists. Additionally, some key advantages of the approach could be highlighted further. For example, this mechanistic model can estimate parameters that are difficult to obtain with statistical modelling such as species distribution models (e.g., biomass, respiration), especially in data-limited environments such as the deep sea. Statistical models often require extensive datasets and in the case of deep-sea benthic species they are typically limited to presence/absence or abundance data, with very few exceptions, such as the models used for validating the present study. Emphasizing that this model integrates physiological and environmental information without requiring large-scale sampling, except for validation, would strengthen the paper’s contribution and practical significance.
Specific comments
Line 22: A comma after species would help the flow of this sentence.
Lines 32-33: Consider moving this sentence up, as it presents the main finding. The authors can then note that it aligns with previous model predictions.
Lines 86-92: These lines focus mostly on POC concentration, but the connection to stressors such as acidification or temperature is not straightforward. I suggest adding 1–2 sentences emphasizing the importance of feeding in CWC ecology (e.g., ability to rapidly exploit food pulses whenever they occur, occurrence of seasonal cycles of growth and reproduction that correlate with food availability), and conclude by noting that feeding has been shown to influence responses to climate stressors (e.g., Büscher et al., 2017).
J. V. Büscher, A. U. Form, U. Riebesell, Frontiers in Marine Science. 4, 101 (2017).
Lines 113-115: I assume that the strong regional contrasts refer to heterogeneity in geomorphology/terrain. Please rephrase.Lines 129: I suggest highlighting that these variables are rarely predicted by statistical models, in the discussion this is an important contribution.Lines 193-195: I think that an "a" should be added to the reference ( Soetaert et al. 2016a). In the work by Soetaert et al, it is mentioned that POC is influenced by passive sinking, hydrodynamic transport, and biological degradation. The results of this study highlight downwelling events often transferring POC to mounds. In these lines, the authors mention advective and passive transport, which may lead non-technical readers to think that upwelling/downwelling were not considered. Please clarify this for non-specialists.
Line 243: The parameter mCWC does not seem to be standardized to coral size/weight or area. Confirm whether it should be expressed per m².Lines 329-332: Can the authors provide the rationale for these decisions (e.g. dividing initial biomass by 3, coral growth/decline enhancement factor of 12)?
Line 396: Should this be “local” rather than “regional”? How easily could the model be scaled up to a larger area?
Line 403: “shows” instead of “show.”
Line 412-413: This sentence is a bit confusing. It seems to refer to environmental factors that were included in the model, but that long-term variations in these factors where not considered, potentially leading to discrepancies between observed and actual CWC cover. The reference to respiration may confuse the reader, please rephrase.
Line 237-239: Considering that the simulation was run for a month, how was benthic respiration extrapolated to the whole year? Did the authors assume that POC is steady throughout the year? I also suggest dividing this sentence in two, to improve flow.
Line 239: The reference to 5,763-9,260 tones is not standardized to area, was the referenced study at exactly the same domain?
Line 440: Specify whether this reference refers to “a” or “b.”
Lines 572–575: On a first read, I found the first sentence slightly contradictory to Fig. A and the fact that this study used a POC model which has highlighted the interaction between tidal currents and CWC-formed mounds in the past (Soetaert et al. 2016a). In the second sentence I realized that this part refers to bottom currents and processes that likely occur in smaller scales. I suggest authors to rephrase these two sentences to avoid confusion.
Figures
Figure 1: A: Adding labels on land shapes will help readers that are not familiar with the area. A short note on the relationship between parent grid, child grid and model domain would also help. These are mentioned in the text, but their relationship is a bit unclear.
Figure 3: Please add latitude and longitude labels.
Figure 7B: Consider comparing modelled versus observed depths directly in panel B for easier interpretation.
Figure 9: For panel A, I suggest using unfilled rectangles to keep the underlying data visible. For panel C, I believe that a simple correlation plot between measured and modelled values (keeping the colour palette for sand and coral areas) would be more informative.
Figure 10: Consider overlaying panel E with panels A and C as an inset, rather than showing it separately. Similarly, clarify the locations of panels A and B. Standardizing the format for all figures that refer to the model area (e.g. Figure 1C, 3, 4A,B), with similar contour line style, colour palettes etc. might help the reader follow the figures easier.
Citation: https://doi.org/10.5194/egusphere-2025-3385-RC3
Data sets
ATLAS Deliverable 2.5: Model code for the Rockall Bank case study area Evert de Froe, Christian Mohn, Karline Soetaert, Dick van Oevelen https://doi.org/10.5281/zenodo.4250150
Model code and software
ATLAS Deliverable 2.5: Model code for the Rockall Bank case study area Evert de Froe, Christian Mohn, Karline Soetaert, Dick van Oevelen https://doi.org/10.5281/zenodo.4250150
Video supplement
Supplemental videos to PhD thesis Evert de Froe: Dinner's Served in the Deep Sea Evert de Froe https://doi.org/10.5281/zenodo.7510506
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 1,459 | 44 | 20 | 1,523 | 42 | 56 |
- HTML: 1,459
- PDF: 44
- XML: 20
- Total: 1,523
- BibTeX: 42
- EndNote: 56
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
OS manuscript egusphere-2025-3385
Reviewer Comments
General comments
Scientific significance
De Froe and co-workers use a predictive mechanistic model to estimate cold- water coral biomass distribution and respiration. Their model successfully reproduces observed reef biomass and respiration patterns. The advantage of this approach can be used to obtain the effect of changing environmental conditions such as ocean temperature, export production or ocean currents.
The study presents a first mechanistic model predicting cold-water coral biomass distribution based on organic matter transport and hydrodynamics. The authors set up the model by cleverly coupling three models, offering a new perspective on the mechanisms driving coral distribution. They demonstrate that coupling organic mater uptake with the cold-water coral model is key to predicting the spatial distribution of these corals.
The authors clearly identify existing gaps in the field and present a study that brings new knowledge and tools that can be applied in future research.
Scientific quality: yes, excellent.
Presentation quality: The manuscript is clearly written and well-structured. The number of figures, conceptual diagrams, and tables is appropriate, and they are of high quality. The supplemental material is warranted and adds value. The authors also discuss the limitations of their work, and the conclusions are well-supported and justified.
I am very positive towards the study, as the findings are important.
Specific comments and detail points:
Line 87 : “This feedback between organisms and their environment can greatly affect how they respond to environmental changes: by modifying their own environment, organisms can rearrange their spatial patterns in response to climate change thereby avoiding a tipping point towards extinction (Rietkerk et al., 2021)” could be reworded. The temporal aspects are not fully resolved within the time scale of the study. Also add a reference relevant to corals.
References in introduction and discussion : Consider reducing to 3-4, as some currently contain five. This will improve readability and focus.
Line 163: replace “ numerous associated animals” with “numerous associated organisms”.
Lines 508-510: The authors could be more specific in this section to enhance clarity and strengthen the argument.
Lines 581 to 585: This section needs to be better explained, particularly the last sentence. I recommend splitting it into two sentences for improved clarity and flow.
References : There are some inconsistencies in formatting within the reference list. Please revise to ensure uniform style throughout.