the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Benthos as a key driver of morphological change in coastal regions
Peter Arlinghaus
Corinna Schrum
Ingrid Kröncke
Download
- Final revised paper (published on 22 Apr 2024)
- Supplement to the final revised paper
- Preprint (discussion started on 12 Jul 2023)
- Supplement to the preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1303', Matthew Hiatt, 16 Aug 2023
Review of “Benthos as a key driver of morphological change in coastal regions”
Review by: Matthew Hiatt
Summary: This manuscript aims to use a hydro-eco-morphodynamic model including the effects of benthos to explain the morphodynamic development of a portion of Jade Bay (German Wadden Sea). The manuscript aims to advance the capabilities of such models in accounting for the effects of benthos at large spatial scales and over multi-year to decadal time scales. The results indicate that the inclusion of bioturbators had the most significant impact on the morphodynamic trajectory of the bay, and the inclusion of all functional groups increased performance. However, the inclusion of increased complexity through parameterization of seasonality in the functional groups actually decreased performance, highlighting the importance of data collection and the tradeoffs between model complexity and performance.
Assessment: Overall, this is a strong, innovative modeling study that presents some very nice results in a most clear and convincing way. While I have several comments that I believe will help increase the clarity of the model setup and results presentation, overall, I think the paper quite good. I would recommend the authors address the following comments with a minor revision. Mostly my confusion was in the description of the different model runs (i.e., Table 3) and how the description of storms (which seemed to have significant impact) were not explained in detail.
Specific Comments:
Line 23: It’s not clear to this reviewer what is meant by “the opposite effect.” Because the morphological change described in the previous line is only given as a general description, one can’t really ascertain what the opposite of that is. I’d suggest a simple change to “…alone would lead to a different pattern than observed.”
Lines 99 101: Instead of “ca.” can it be substituted for clarity by “approximately” or “about” ? I see that ca. is used throughout. I suppose that’s fine; it’s just less common and clear (I had to look it up).
Line 137: I’m unclear about the line “the latter three were extracted from the hydrodynamic model results.” Does this mean that the input for the model including the benthos parameterizations were determined from model runs without benthos? If so, that’s fine but needs to be stated as such. However, it does beg the question that if benthos is such a huge contributor to morphodynamics as the paper finds, then how can parameters like bed shear stress and mud content be determined without benthos in the first place? This is a non-issue if the topography used in the simulation to gather inundation time, shear stress, and mud content is representative of current conditions (although it seems to me mud content would need a validation). It’s sort of an intractable problem, but perhaps the authors can comment and make this statement a little clearer as to how these parameters are pulled from the model (and which model run they’re pulled from).
Line 209: What is meant by “vegetation proof?” I’ve not heard that term before.
I do not understand Table 3 at all. The entries like Des0, sta0, etc. can be surmised, but it’s not clear. I’d recommend an overhaul of this table for clarity.
Lines 256-258: This is the first mention of storms. How are storms defined? I think we need more information here as to how storms are implemented in the model and how the sensitivity testing is implemented. Do runs with no storms mean wind is turned off? As written, I’m not really clear on what to expect from the model runs, especially given my lack of understanding of table 3 (Previous comment).
Lines 306-320: These lines greatly increased my understanding of the sensitivity testing that was done by turning parameterizations on and off. It was very clearly written. It would have been helpful for something similar a little early alongside an improved Table 3.
Lines 321-328: These lines indicate what was meant by the “opposite effect” in the abstract (regarding my first comment). I suggest using the language here, which is very clear, to help the writing in the abstract.
Figure 6b – The y axis reads average main channel depth and the caption reads average depth change. Please adjust for consistency between the two to improve clarity. I assume it’s supposed to be the change so it aligns with Figure 5b.
Figure 7b: What is the greenish grey area?
Figure 7: I am somewhat unclear on this Figure. So the colors represent the number of flat types that were changed in a given region? If that’s the case, I get it, but am still wondering what is the purpose of this information? It’s not abundantly clear to mean and is an opportunity for the manuscript to become clearer in purpose. In essence, why does it matter that more than one flat type changed? I would expect it’d be more useful to understand the trajectory of each type of flat. In other words, the transition from one flat type to another or the change in composition (like Figure 9).
Technical Edits:
Line 129: fix “spacesince”
Figure 7 caption: NLWKN is not known to the reader.
Line 500: some sort of typo here: “Especially an import of mud into the bay is…” Maybe “an” should be removed?
Citation: https://doi.org/10.5194/egusphere-2023-1303-RC1 -
AC1: 'Reply on RC1', Peter Arlinghaus, 21 Sep 2023
Reply on RC1
We thank Referee #1 for the constructive comments that have helped us to clarify and improve our manuscript. Our responses to the specific questions/requests (in bold & italic) are listed below.
Assessment:
(…)Mostly my confusion was in the description of the different model runs (i.e., Table 3) and how the description of storms (which seemed to have significant impact) were not explained in detail. (…)
Thank you for pointing this out. We have improved the description in naming the functional groups and parameter settings in the associated sensitivity model runs.Abbreviations for the functional groups, seasonality, hydrodynamic forcing and sediment parameter are:
Biomixers = mix
Accumulators = acc
Stabilizers = sta
Seagrass = graInclusion of all four functional groups = all
Abiotic model run without consideration of any benthos effect = abio
Seasonal variation of benthos excluded/included = no / abbreviation of specific functional group(s)
Hydrodynamic forcing excluding/including storm surges = T / TS
Erosion rate by default / scaled by factor of 10 = 1 / 10
The experiments are named by combination of the different model features separated by an underscore and read as:
Modeled functional groups_Seasonality_Hydrodynamics_Erosion Rate
A full set of experiments with the modified names is provided in the updated Table 3, includingall_no_T_1
all_no_TS_1
all_mix_TS_1
all_all_TS_1mix_no_T_1
mix_no_TS_1
mix_mix_TS_1sta_no_T_1
sta_no_TS_1
sta_sta_TS_1
sta_no_T_10
sta_no_TS_10
sta_sta_TS_10acc_no_T_1
acc_no_TS_1
acc_acc_TS_1
acc_no_T_10
acc_no_TS_10
acc_acc_Ts_10gra_no_T_1
gra_no_TS_1
gra_no_T_10
gra_no_TS_10abio_no_T_1
abio_no_TS_1
abio_no_T_10
abio_no_TS_10
A short description on how storms were implemented was shown in the supplementary material. We will expand this paragraph with a more detailed description.Specific comments:
Line 23: We agree with the reviewer. This line will be changed in order to emphasize that “opposite” refers to morphological changes in terms of erosion and deposition.
Lines 99 101: We will change ca. to approximately to prevent confusion.
Line 137: Indeed, some of the parameters were determined from model simulations, that did not contain benthos. However, there is no conflict in using those parameters for predicting a current species distribution. The reviewer specifically questions if mud content and shear stress can be determined without benthos. The term shear stress should not be mistaken for critical shear stress for erosion. The latter is indeed strongly impacted by benthos, but the former is primarily depending on the hydrodynamics at the sediment-water interface. For the prediction of species distribution, bottom shear stress but not critical shear stress for erosion were used.
The applied mud contents are based on measurements (not simulations) and was used as a proxy for estimating species distribution. It is correct that benthos strongly impacts local mud content (and vice versa), but the induced small-scale changes may accumulate on time scales of years to decades to generate a large-scale effect. We agree that a more adaptive species distribution, which includes the mutual and dynamic feedback between mud content and benthos would be more accurate. This was however, out of scope in this study. We assume that other environmental parameters which were based on simulations and used for the estimation of species distribution such as inundation time or salinity are not significantly influenced by benthos.Line 209: Instead of “vegetation proof”, we will use the term “vegetation cover”
Table. 3: Please see our response to comment 1.
Lines 256-258: Actually storms were mentioned earlier in chapter Model setup for the study area in line 230. However we do understand why it was easy for the reviewer to overlook this point. In line 230 we only mention storms once and point to the supplementary. We will add another brief explanation here on how storms are implemented and extend the description in the supplementary.
Lines 306-320: We agree and will add a remark earlier in the text.
Lines 321-328: We agree and will update the abstract accordingly.
Figure 6b. This is true, caption and axis are inconsistent. We will improve this figure to provide more precise and quantitative information. The caption inconsistency will also be corrected.
Figure 7b: Plot 7b is not our work, but was adopted from Ritzmann and Baumberg (2009) with permission (as indicated in the caption). This plot is based on measurements and the “greenish grey area” indicates the area where no measurements were made. We will add this information to the caption.
Figure 7: One of the main purposes of this study is to show that simulation results are significantly improved if benthos impact is added. In order to prove this, we used historical morphological data, and data indicating the sediment change for model assessment. The sediment change data are shown in figure 7b. These data and this plot were not created by us, but adopted from Ritzmann and Baumberg (2009) with permission. The plot depicts the sediment change in terms of flat type change. Unfortunately we were not able to get the raw data that were used to produce the plot. We agree with the reviewer, that it would be more useful to compare composition changes and temporal transitions maps from one flat type to another. Because of lack of field data we had to stick to the flat type change comparison.
In order to avoid confusion here, we will provide explanation in the text on why flat type comparison was chosen, rather than actual composition change.Technical Edits:
Line 129: Will be changed as suggested by the reviewer.
Figure 7 caption: NLWKN (“Niedersächsischer Landesbetrieb für Wasserwirtschaft, Küsten- und Naturschutz”) is state agency that is subordinate to the German state administration. We will explain this in the text.
Line 500: Will be changed as suggested by the reviewer.
Citation: https://doi.org/10.5194/egusphere-2023-1303-AC1
-
AC1: 'Reply on RC1', Peter Arlinghaus, 21 Sep 2023
-
RC2: 'Comment on egusphere-2023-1303', Anonymous Referee #2, 18 Aug 2023
Review of “Benthos as a key driver of morphological change in coastal regions”
In this manuscript, the authors present an innovative model that accounts for different functional groups or organisms to affect morphodyanmics. Through expansion of the SCHISM modeling framework, the authors incorporate bioturbators, biostabilizer, and seagrasses and demonstrate that bioturbators have significant impacts on the morphology of their entire domain. This work is innovative and novel, coupling this advance in numerical modeling with the use of machine learning techniques for species abundance modeling and provides insight into the importance of biota in estuarine dynamics. I particularly liked the discussion about model complexity as well. However, while I feel that the modeling work done is of high quality, parts of the manuscript lack clarity and there is a mixing-and-matching of model scenarios that make it difficult to assess the relative roles of the different biological groups. This novel work is certainly worth publication and I was excited to read the manuscript. I have some suggestions for clarity and for defending certain choices made that would strengthen the paper. I look forward to seeing a revised version!
Specific Strengths:
- Line 43: good “needs” statement
- Line 61: explanations for why still at exploratory phase rather than explanatory
- Fig 7: cool idea
- Great finding about the distal hydrodynamic effect of seagrass, even when affecting sediment transport and morphodynamics on local level
- Line 535: really nice overarching / generalizable finding
- Section 5.3 is well-written and convincing
Major Comments:
- Presentation of Model Runs. I believe that you would strengthen your paper by changing how you label and compare model runs. Firstly, the naming conventions (as those presented in Table 3) are confusing for a casual reader, and these same conventions are carried on throughout the paper. I had to have Table 3 next to me during the entire read to keep the model runs straight. Furthermore, in later analysis, you compare model runs intending to show the impact of biota, but some of the runs have storms and some do not (i.e. Figure 8). While you chose the model runs to optimize RMSE, it is not possible to evaluate the effects of the different biological groups if the baseline hydrodynamics are not the same.
- Unintended consequences of results. While this is not a problem with your analysis/modelling, but one takeaway from the paper is that seagrass didn’t do much to affect the morphology. I like that in the supplementary you show that seagrass does affect the hydrodynamics. However, I worry that readers (who would cite your paper) may use your results to say that in general seagrasses don’t affect morphodynamics, which is not always true. I recommend making sure that is clear in your discussion by citing a few papers that show in other bay geometries seagrasses do make a different in sedimentation.
- Additional representation of quantitative results. Need additional results figures besides just maps. The maps only qualitatively show differences, no way for reader to quantitatively compare. Furthermore, the maps become repetitive and have opportunity to communicate results in different and more interesting ways
Some Ideas:
- Graph distinguishing between accuracy of channel and tidal flat elevation change, or benthos effect on channel vs tidal flat elevation/morphology
- Graph with the distance that effects from benthos communities extend across domain
- Graph showing amount of interaction b/w functional groups (calculate RSME for smaller domains that have similar [read: normalized] metabolic intensities of the different species, then compare distal vs proximal effects and channel vs flat effects
- Explanation of model choices. Please elaborate on why you made certain modeling decisions by either citing more literature or discuss the decisions. For example, elaborate on why you increased the erosion factor by 10, and why there was no combination of E10+Storm+Season2. Additionally, did SSC remain constant for the study period (in real life)? There is potentially big seasonality and interannual variability. Do you have any data to justify this choice beyond the one piece of data that you got 40 mg/L from? Additionally, do you assume that benthos abundance numbers remain constant? You mention (line 473) that bioturbator dominance has increased, but it seems that the model only uses numbers based on 2009?
- Consistency in naming. Please improve the consistency with how you refer to benthos functional groups/species. Sometimes call them by species name, sometimes by functional group, and sometimes by their morphological effect (e.g., “destabilization,” “accumulation” as in Fig 8)
- Are your results dependent on the initial distribution of biota? You could consider calculating the RMSE for areas with high biomass of the given species (for example seagrass) to get at the local effects and normalize for the fact that seagrass is only found in a small area compared to more extensive presence of bioturbators. Right now, it seems that you Unintentionally indicating that bioturbators are the most important to include as benthos in eco-morphodynamic models, but results are skewed that way because of high abundance throughout domain compared to patchy abundance for other functional groups.
- Strengthen answer to research question 2. As is written now, your rational for choosing the model runs you show is not compelling.
- Can you isolate the individual contributions of each functional group within All2 run? That way you take the most accurate run and assess individual contributions. However, I understand this may be difficult to do given feedbacks in the model.
- Also the approach you take is slanting your results towards bioturbators because they are the most abundant. Can you normalize by downsizing to a subdomain for this analysis that has similar/same hydrodynamic conditions and similar biomass/abundance/metabolic intensity for each functional group?
- If unable do 1 & 2 based on model limitations, choose same hydrodynamic conditions. Not a strong argument that you choose to compare the four runs based on how low their individual RMSE values are because wouldn’t expect ones with small biomass/abundance to accurately match the measured morphologic change
Figures:
In general, I would suggest switching colors so erosion is associated with the color red. Also, I would consider using a different colorbar to make it easier for readers to see the more subtle changes.
Figure 1: This figure would benefit from a scale bar (I know latitude and longitude are shown, but it would be clearer) and a north arrow.
Figure 2: Either in the figure and/or in the text, it may help to provide a common name for the species (i.e. clam, cockle, snail) and/or add a cartoon of the organism. While some of your readers will be familiar with these species names, most geomorphic modelers, who will like this work, will not be familiar with them. I would also specify in the figure and in the text that biomass stays constant throughout the model runs and that there are not feedbacks that allow biological communities to shift.
Figure 3: Panel A is nicely shown. I am not sure that the Taylor diagram is necessary, but if you like it that’s okay. If you keep it, I recommend changing the location symbols to two different shapes so that you can read it without color.
Figure 4: Looks fine. Currently, you don’t really use the colors in other parts of the paper – you could continue use of these colors to help identify model runs throughout. Just a thought.
Figure 5: I suggest that instead of using model run names on the figure, describe what they are to give the reader a more intuitive understanding rather than having to refer to Table 3 (for example, instead of writing “All2”, you could write “All Benthos + Storms +Seasonality”. I also think it would be best to have the measurements as panel A, since that is what you are comparing the other panels to. You should also indicate on the figure which ones have a heightened erosion rate directly (Ref1b). I was wondering why you use All2 when you state in the paper that seasonality doesn’t improve fit and panels c and d do not have seasonality (I suppose because they are abiotic?). However, that is a bit confusing and would suggest using All1 instead of All2 for better comparison. I would also recommend flipping panels C and D, because they way you present the story, you say that the reference case doesn’t do a good job (Figure 5d currently) unless you increase erosion rate by a factor of 10 (Figure 5c currently). It would tell the story better to present them in order.
Figure 6: I like this figure with the exception that Ref1 and Ref1b should be switched, and I would maybe write what the simulations represent rather than the run name.
Figure 7: Again, the measurements should go as panel A, since the other panels refer to it for change. I would also again suggest reordering and changing to All1 instead of All2. If you don’t switch, why did you chose to use All2? I also think I understand the roman numerals (flat type), but could you explain better in the caption? Please define what they grey area is.
Figure 8: I wrote about this more above, but it doesn’t make sense to compare runs with different hydrodynamic conditions (storms vs. no storms), even if they have higher RMSE. I suggest compared runs with the same background forcings. Also consider writing in the corner of each panel the total amount of erosion and deposition (sum of red and sum of blue) to give the reader a quantitative way of comparing the simulations instead of solely looking that the colors.
Figure 9: Same comments as figure 8. I like that you wrote the model run subtraction on the figure itself, but it would be even better to use a more descriptive name compared to the abbreviation.
Figure 10: I love this figure. I would maybe suggest removing sediment particles since they are difficult to see and they always go with arrows. I would also say that this doesn’t represent a generic tidal embayment, because the results very much depend on the distribution of biota. It still is a great conceptual diagram for Jade Bay. Please label the biofilms in panel d, and consider using other colors than blue or purple for the arrows as they are hard to see.
Line-by-line edits:
line 110 would be a good time to reiterate or go into more specifics, still reads as vague
Line 120: reference the figure (fig. 2b-f) or restate this in figure caption because reader may have forgotten that modeled biomass based on abundance
Line 124: clarify what type of species they are (common names?)
Line 161: Explain how to calculate a and b.
Move paragraph starting at line 288 above previous paragraph.
Line 321: Offer a brief explanation of how to calculate cosine similarity
Line 464-466: Please give more (process-based) explanation for why bioturbators lead to enhanced resistance to erosion
Line 85: “drowned”Line 93: should be “1. To what extent do benthos”
Line 157: “formulae” (?)
Line 365, 369, etc.: change all instances of “despite of” to just “despite”
Fig 7b: “Measurements”
Line 541: “growing concern about whether”
Line 545 and 551: “in the future”
Line 552-553: “are imperative for exploring…”
Citation: https://doi.org/10.5194/egusphere-2023-1303-RC2 -
AC2: 'Reply on RC2', Peter Arlinghaus, 21 Sep 2023
Reply on RC2
We thank Referee #2 for the constructive comments that have helped us to clarify and improve our manuscript. Our responses to the specific questions/requests (in bold & italic) are listed below.
- “(…) the naming conventions (as those presented in Table 3) are confusing for a casual reader.”
Thank you for pointing this out. We have improved the description in naming the functional groups and parameter settings in the associated sensitivity model runs.Abbreviations for the functional groups, seasonality, hydrodynamic forcing and sediment parameter are:
Biomixers = mix
Accumulators = acc3
Stabilizers = sta
Seagrass = graInclusion of all four functional groups = all
Abiotic model run without consideration of any benthos effect = abio
Seasonal variation of benthos excluded/included = no / abbreviation of specific functional group(s)
Hydrodynamic forcing excluding/including storm surges = T / TS
Hydrodynamic erosion rate by default / increased by factor of 10 = 1 / 10
The experiments are named by combination of the different model features separated by an underscore and read as:
Modeled functional groups_Seasonality_Hydrodynamics_Erosion Rate
A full set of experiments with the modified names is provided in the updated Table 3, includingall_no_T_1
all_no_TS_1
all_mix_TS_1
all_all_TS_1mix_no_T_1
mix_no_TS_1
mix_mix_TS_1sta_no_T_1
sta_no_TS_1
sta_sta_TS_1
sta_no_T_10
sta_no_TS_10
sta_sta_TS_10acc_no_T_1
acc_no_TS_1
acc_acc_TS_1
acc_no_T_10
acc_no_TS_10
acc_acc_Ts_10gra_no_T_1
gra_no_TS_1
gra_no_T_10
gra_no_TS_10abio_no_T_1
abio_no_TS_1
abio_no_T_10
abio_no_TS_10
A short description on how storms were implemented was shown in the supplementary material. We will expand this paragraph with a more detailed description.- “While you chose the model runs to optimize RMSE, it is not possible to evaluate the effects of the different biological groups if the baseline hydrodynamics are not the same.”
We agree, that the baseline hydrodynamics should be the same. This will be updated in the plots in the revised manuscript. Nevertheless, the main message and interpretations are not affected.
- “However, I worry that readers (who would cite your paper) may use your results to say that in general seagrasses don’t affect morphodynamics, which is not always true.”
Indeed. This is a misleading message which our study would like to avoid. Despite that seagrass impact is less striking when compared to the major impact by bioturbators in the Jade Bay, seagrass meadows are indeed impactful and can change the morphology by several centimeters and up to tens of cm within a few years. More important is that seagrass meadows can modify the hydrodynamics and thus morphodynamics at scales well beyond their habitat. The discussion will be updated in this regard, highlighting that the differences in impact of different functional groups is strongly depending on their distribution and abundance/biomass, which can vary strongly among different sites. We will also add more references in the discussion highlighting the impact of seagrass.- “Need additional results figures besides just maps. The maps only qualitatively show differences, no way for reader to quantitatively compare.”
We will add quantitative comparisons on the benthos effect in tidal channels and tidal flats. The newly added figure will show the average depth change in the channel as function of the distance from the tidal inlet for different biotic and abiotic runs and the depth change at the tidal flats as function of distance from the tidal inlet. It will be clear from this figure that, according to the measurements, the main channel have accumulated abundant sediment during the investigated time period, which was reproduced by the simulations including biomixers while other simulations failed.
The new figure and interpretation will be added to the main text.
The other two suggestions from the reviewer to create new maps were not realized for the following reasons:
A graph with the distance that effects from benthos extends across domains is in principal possible to create. In case of the seagrass this can be done since it is very locally distributed. For other species however, this is not feasible since they are spread over the whole domain. And only considering areas with high biomass may lead to wrong results since benthos impact in region of small benthic biomass might be higher than in regions of high biomass depending on hydrodynamics (Cozzoli, 2016).
The third suggestion is about comparison between distal and proximal effects. This suggestion is similar to the previous one. If we understand the reviewer right, he/she asks for normalization of benthos impact of morphology, normalized by the metabolic rate of the different species. This is indeed a very interesting suggestion, which however we can unfortunately not realize. The reason is that for stabilizers and for seagrass no metabolic rates are known. Their abundance relies on rough estimates for stabilizers (Widdows and Brinsley, 2002; Daggers et al., 2020) and 2D abundance maps for seagrass (Adolph, 2010). Generating estimates for the metabolic rate are possible but will come with huge uncertainties. Furthermore it is not clear if the impact of macrophytes and micropythobenthos scales similar to the relation between microbenthic fauna and their metabolic rate since they are autotrophs and partially immobile.
- “Please elaborate on why you made certain modeling decisions by either citing more literature or discuss the decisions. (…) Additionally, did SSC remain constant for the study period (in real life)? (…) Additionally, do you assume that benthos abundance numbers remain constant? (…)”
One of the ideas of this paper is to test the impact of different biological, hydrodynamic and sedimentological properties on the model outcomes in order to estimate the sensitivity of the morphology. For this reason major benthic functional groups including seasonal changes, storm impacts and sediment properties, namely the erosion rate, were investigated. In setups that included bioturbators (biomixers) the suspended sediment concentration (SSC) is strongly increased. Higher SSC makes the effect of filter feeders more pronounced since there is more material to deposit. Thus, making simulations with only accumulators is not sufficient to disentangle their impact in a realistic (all benthos) scenario, since their impact will be much weaker compared to those in a combined simulation of species. Increasing the hydrodynamic erosion rate by a factor of 10 aims to evaluate the effect that accumulators are having in the all benthos run.
The SSC coming from the open boundary is constant in the model. Likely there will be seasonal variability in SSC which we did not implement due to the lack of measurement data. To our knowledge there is no existing dataset showing turbidity values over longer time spans and on larger spatial scales in Jade Bay. The measurements we were able to find typically cover one or a few points measured over one or a few tidal cycles (Götschenberg and Kahlfeld, 2008; Becker, 2011). Measured values are in the same range as the one chosen in the paper (Becker, 2011). Other modeling studies from the Jade Bay show comparable amounts of suspended sediment as found in our simulation of Jade Bay (Kahlfeld and Schüttrumpf, 2006). This will be emphasized by putting a plot into the supplementary, showing the average SSC in the Jade Bay. Explanations and mentioned citations will be added to the main text.
There is unfortunately no field measurements on the inter-annual variability of benthos in the study area. We adopted the distribution derived from 2009 measurement. Seasonal variation was implemented as a simple function in some of the experiments. Other experiments assumed a stationary distribution. As pointed out in Arlinghaus et al. (2021) and Arlinghaus (2023, PhD thesis in press), the lack of species abundance/biomass distribution data is one of the major problems hindering the development of large-scale morphodynamic modeling. This can/should be improved by implementing regular monitoring schemes. We would like to emphasize here that, although the field dataset lacks a temporal evolution, it represents the most comprehensive dataset from the Wadden sea that is used in a large-scale bio-morphodynamic modeling study, see Table 2 in Arlinghaus et al. (2021).
We will add discussion of each point above in the main text.- Consistency in naming. Please improve the consistency with how you refer to benthos functional groups/species. (…)
We agree. This has been clarified in the response to the first major comment. Please see above. - You could consider calculating the RMSE for areas with high biomass of the given species (for example seagrass) to get at the local effects and normalize for the fact that seagrass is only found in a small area compared to more extensive presence of bioturbators.
This comment is similar to an earlier one (comment 4). Unfortunately, as explained above, the normalization regarding biomass of metabolic rate for seagrass and other functional groups is not feasible with the data that we have due to large uncertainty. - Strengthen answer to research question 2. As is written now, your rational for choosing the model runs you show is not compelling.
We agree that especially the question for combined impacts of benthos is not elaborated thoroughly enough. We will do so in the text.
- Can you isolate the individual contributions of each functional group within All2 run? (…)
As assumed by the reviewer this is only partially possible since the functional groups interact and thus a single functional group simulation can never completely represent the effect that this functional group would have when presented with other functional groups. However, one way interaction is mediated between functional groups, which is due to the available amount of suspended sediment. To reach a comparable level of suspended sediment to the scenarios in which the sediment is strongly destabilized by bioturbators, the hydrodynamic erosion rate was artificially increased by a factor of 10. A respective note will be added to the text.
- “Also the approach you take is slanting your results towards bioturbators because they are the most abundant. Can you normalize by downsizing to a subdomain (…)”
As explained in several comments above this is not feasible since biomass values for seagrass and stabilizers are not known. To address this we will add a discussion in the main text, pointing out the need to further investigate the relative impact of species/functional groups, compared to its biomass and metabolic rate.
- “If unable do 1 & 2 based on model limitations, choose same hydrodynamic conditions. (…)”
In context of suggestion 1 it is valid to use RMSE to compare the functional groups, even though they are presented in different abundances and biomasses. Because it shows, which functional group contributes most to the changes observed in the study area.
If we understood the reviewer right, he/she wants to get answer for the following question: “Assuming the same biomass and/or metabolic rate and the same spatial distribution in a patch that experiences the same hydrodynamic conditions, which functional group will have the largest impact on the morphology? If this cannot be answered, cross out biomass and metabolic rate from this question”.
As mentioned above, answering this question offers sufficient material to make a new study by itself. For example, to answer this question we would need to make a new model simulation, choosing a small patch where all four functional groups are abundant and on the other hand remove all benthos outside this patch (since benthos effect reaches well beyond their habitats). We also need to be careful when choosing the patch, since every part of the bay has a unique impact on the overall hydro-morphodynamics of the whole system. We believe that answering this question is very interesting and worthwhile, but would go beyond the scope of this study and lead to a very lengthy but not well streamlined paper. We hope the reviewer would agree with our argument.
Figures:
In general:
It is not uncommon to use blue color for erosion and red for accumulation of sediments as can be seen in Benninghoff et al., (2019) or Brückner et al., (2021). For this reason we prefer to keep the color presentation.
In the choice of the colormap we are limited to a divergent colormap. In order to better distinguish the subtle changes, seismic instead of bwr colormap will be used in an updated plot, which has higher saturation in color. Additionally the range for the depth change will be adapted in order to emphasize small changes. The information lost, when cropping the colorbar will be compensated by the new picture suggested by the reviewer and that will be added to the updated version of the paper. It shows the actual depth change in the channel and tidal flats.Figure1:
We agree: Scale bar and North arrow will be added to the plot.Figure 2:
Both common names and cartoons will be added to the figures. That the biomass is constant will be mentioned in the text.Figure 3:
We will keep the Taylor diagram. One of the circles will be changed to a triangle as suggested.Figure 4:
This is a good idea. We will color all model run names that appear in figures.Figure 5:
The figures will be updated according to the new names. Also increased erosion rate is indicated in the new model run names.
As can be seen from figure 4, adding seasonality of bioturbators (biomixers) improves simulation results, while adding seasonality of filter feeders into the simulation decreases the model performance. This is why All2 (now: all_bio_TS_1) was presented.
Indeed, the abiotic runs have no seasonality of benthos because they do not include benthos. We think it is valid to keep All2 since seasonality is exclusively linked to benthos.The order of panels a-d will be changed according to the suggestion.
Figure 6:
The figure will be updated by using the new names of the experiment according to the suggestion.Figure 7:
Panels will be reordered as suggested. All2 (now: all_bio_TS_1) is chosen because it has the least RMSE. In figure 7 and 5, the aim is to compare the measurements to the best simulation (least RMSE) that could be achieved. This does not conflict with having no seasonality of benthos implemented in Ref1b and Ref1, because these runs are abiotic (see response to Figure 5). On the other hand, since we are comparing to measurements here, that inherit benthos seasonality, it makes sense to compare with a simulation which includes benthos seasonality.
The roman numbers indicating the five areas with pronounced changes are chosen to compare the measurements with the simulations. An indication of that will be added to the caption of the figure.Figure 8:
We agree that it makes no sense to compare runs with different forcing. Thus the compared runs will be adjusted.
Indicating the amount of erosion and deposition in the plot is a very good idea. It will be added to the figure.Figure 9:
The reviewer’s suggestions will be adopted to update the figure.Figure 10:
We are happy about the reviewers appreciation of this figure! The figure will be further improved according to the reviewer’s suggestion.Line-by-line edits
line 110: We will extend the description accordingly.
Line 120: The reference for figure 2b-f will be added.
Line 124: Common names will be added:
Line 161: How a and b are calculated is already described in the lines below and stated in the equations (5) and (6). However, we found a typo in equation (5). It is corrected as:
Move paragraph starting at line 288 above previous paragraph: Order of paragraphs will be switched
Line 321: A brief explanation on how to calculate the cosine similarity will be given.
Line 464-466: A further explanation will be added to the main text.
Line 85: Correction will be applied.
Line 93: Correction will be applied.
Line 157: Correction will be applied.
Line 365, 369, etc.: Correction will be applied.
Fig 7b: Indeed we meant “plot” and not measurements. This will be clarified in the caption of figure 7.
Line 541: Correction will be applied.
Line 545 and 551: Correction will be applied.
Line 552-553: Correction will be applied.
Refernces used in the response letter:
Adolph, W.: Praxistest Monitoring Küste 2008: Seegraskartierung: Gesamtbestandserfassung der eulitoralen Seegrasbestände im Niedersächsischen Wattenmeer und Bewertung nach EU-Wasserrahmenrichtlinie. NLWKN Küstengewässer und Ästuare, (2):1–62, 2010.
Becker, Marius. Suspended Sediment Transport and Fluid Mud Dynamics in Tidal Estuaries. PhD Thesis. University of Bremen, 2011.
Benninghoff, M., Winter C., 2019. Recent morphologic evolution of the German Wadden Sea. Sci. Rep. 9, 9293. https://doi.org/10.1038/s41598-019-45683-1.
Brückner, M., Schwarz, C., Coco, G., Baar, A., Boechat Albernaz, M. and Kleinhans, M.: Benthic species as mud patrol - modelled effects of bioturbators and biofilms on large-scale estuarine mud and morphology. Earth Surf. Process. Landforms. 46: 1128– 1144. https://doi.org/10.1002/esp.5080, 2021.
Daggers, T. D., Herman, P. M., and van der Wal, D.: Seasonal and Spatial Variability in Patchiness of Microphytobenthos on Intertidal Flats From Sentinel-2 Satellite Imagery. Frontiers in Marine Science, 7, 392. doi:10.3389/fmars.2020.00392, 2020.
Götschenberg, Axel; Kahlfeld, Andreas. The Jade. In: Die Küste 74. Heide, Holstein: Boyens. S. 263-274. 2008.
Kahlfeld, Andreas & Schüttrumpf, Holger. (2006). UnTRIM modelling for investigating environmental impacts caused by a new container terminal within the Jade-Weser Estuary, German Bight.
van Maanen, B., Coco, G., and Bryan, K. On the ecogeomorphological feedbacks that control tidal channel network evolution in a sandy mangrove setting. Proc. R. Soc. A. 471, 20150115. doi: 10.1098/rspa.2015.0115, 2015.
Ritzmann, A. and Baumberg, V.: Forschungsbericht 02/2013 – Oberflächensedimente des Jadebusens 2009: Kartierung anhand von Luftbildern und Bodenproben; NLWKN Niedersachsen, 2013.
Widdows, J. and Brinsley, M.: Impact of biotic and abiotic processes on sediment dynamics and the consequence to the structure and functioning of the intertidal zone. Journal of Sea Research, 48, 143-156. doi:10.1016/S1385-1101(02)00148-X, 2002.
Citation: https://doi.org/10.5194/egusphere-2023-1303-AC2