Review of "On the relative role of abiotic and biotic controls on channel network development: insights from scaled tidal flume experiments" by Sarah Hautekiet et al.

Dear Editor and Associate Editor,

the paper describes a large-scale experiment on vegetation effects on the evolution of a tidal network. The experiments were carried out in the Metronome at the Utrecht University. Vegetation seeds were sown by dispersing them or manually creating some patches during the evolution of the system. The experiments were then compared to control experiments where no vegetation was used. Pictures of the tidal system were taken at different stages of the experiments to monitor vegetation growth and the channel geometric properties (with and without vegetation). In the experiments, vegetation favoured the development of a denser and more efficient network.

The issue is timely and of broad interest, particularly in view of the current great concern about the fate of coastal environments threatened as they are by climate change. The paper is well-written and well-conceived, but it would benefit from some minor clarifications and corrections. Thus, I recommend publication with minor revisions as the manuscript meets the kind of broad interest commanded by the readership of your Journal.

Please read below my comments (italics refer to the text of the manuscript).

Questions

• What was the duration of each experiment (days) and what would it correspond to in a real tidal system (years)?
• How did you ensure that the morphodynamic equilibrium was reached? Did you use a local or global criterion? The average difference between two successive DSM was computed and, if the difference was smaller than a certain threshold, the morphodynamic equilibrium was reached? Did the vegetation encroachment increase the speed at which the system reached the morphodynamic equilibrium? Could this trend be in line with previous field investigations in tidal environments?
• What are the limitations of using this vegetation species for large-scale experiments on the morphodynamic evolution of a tidal network? Vegetation has many functions in a tidal environment, for example it increases friction and favours deposition by sediment trapping and organic production. Are all these functions potentially reproduced (or reproducible) in the experiments? Or is there a function that is more dominant than the others in the experiments?
• Does the vegetation grow in a specific elevation range related to the tidal range? Does the vegetation biomass peak at a specific elevation?

Figures

Figures 4-11 and their labels look a bit blurred on my pdf file. This could be due to the pdf file compression but, anyway, please make sure that the overall resolution of each picture is high enough (usually 300 dpi).

Figure 1. On the top left, the panel labelling seems starting with “b)” and it gets a bit confusing when I look at the inset on the top right. I would replace label “b)” on the top left with “a)” and delete the labels in the inset on the top right leaving only the black square.

Figure 2. Could you add photographs where you see the channel network and the vegetation on the floodplains? I am thinking of photographs like those in Figure 5 in Weisscher et al., 2022. It helps to better visualise the experimental setup.

Figures 3, 4, Sup_f01, Sup_f02, Sup_f03, Sup_f04, Sup_f05. Tick values on some vertical axes are missing.

Figure Sup_f05. In the panel referring to hydrochorus vegetation at 3071 cycles, the sea basin is represented with a darker blue. I think it is only a plotting issue, but it could be worth checking it out.

Figure 5. This is a very nice picture. It seems that you have a figure composed of only a single column. Is it done on purpose? If you stretch the horizontal axis as in Figure 8, does it get too distorted? Moreover, tick values on some vertical axes are missing. If you use panel labelling, you can refer to specific panels when describing the figure in Lines 256-269.

References

Lines 606, 625. “/a -n/a” appears in the references Temmerman et al. (2005) and Vandenbruwaene et al. (2011). Is that right?

Lined comments

• Line 27-28. “Tidal channel networks dissect the landscape, which play a key role in the ecosystem functioning since the channels act as major flow paths for water, nutrients, sediments, and biota”. The sentence is grammatically correct, but by reading it quickly, “which” seems referred to “landscape” rather than “tidal networks”. I would suggest replacing “which” with “and” or using a passive version in the main clause such as “Tidal environments are dissected by channel networks that play …”
• Line 121. “…allowing connection of the intertidal basin with the “sea basin”…”. Please use “connecting the intertidal basin with the “sea basin””
• Line 146. “Based on Lokhorst et al. (2019), experience in the Weisscher et al. (2022) experiments and pilot experiments not reported here”. Is there a reference (paper or conference abstract) for the "pilot experiments not reported here" (are the pilot experiments mentioned in line 173?)? I would simplify the sentence as "Based on previous experiments (Lokhorst et al., 2019; Wesscher et al., 2022;, ...), a single plant species was ..."
• Line 147. Please use “a single plant species” instead of “one plant species”
• Line 149. “alfalfa”. Plant species were written in italics in the manuscript. Please write alfalfa in italics.
• Line 150. “it does not establish in unsuitable locations”. Do you mean “grow”?
• Line 196. “… with a resolution of 4000 of 6000 pixels …”. Is it 4K or 6K?
• Lines 198-199. “The images were calibrated for internal and external parameters (i.e., lens correction, geometric rectification) before they were stitched.”. Which calibration method did you use for correcting lens distortion? Checkerboard method by Zhang (1998) or another one? Could you add the reference for the method?

Zhang, Z. (1998). A flexible new technique for camera calibration. Technical Report MSRTR-98-71 Microsoft.

Was the geometric rectification carried out using python built-in functions in OpenCV? If so, please add a reference of the python package (e.g., Bradsky, 2000).

Bradski, G. (2000). The OpenCV Library. Dr. Dobb’s Journal of Software Tools, 120; 122-125.

• Line 164-165. “… (to avoid subsurface drainage) and avoid plant growth in channels …”. Please delete brackets and the repetition of "avoid" as "...to avoid subsurface drainage and plant growth channels..."
• Line 206. “Basic corrections were applied.”. Please give a brief description of the basic corrections applied.
• Lines 215-16. “…along the length direction of the flume and over the complete width of the flume (3 m).” Could it be “along the full length of the tidal basin (… m) and across the complete width of the flume (3 m)”?
• Lines 252-253. “the time scale […] was shifted by up to about 1000-2000 cycles”. "was shifted by" or "was shifted up to" ? Please choose the one that best describe the procedure.
• Line 259. By introducing panel labelling, you could refer to specific panels in figure 5 when describing evolution phases.
• Lines 283-284. “Even though the vegetation cover remained lower than expected, it still affected the channel network characteristics (see section 3.3).”. Since this point is addressed in the following subsection, I would suggest deleting this line here.

I can definitely say I enjoyed reading the paper that is well written, well organized, concise but exhaustive in describing the methodology and in presenting and discussing the results.

Based on scaled flume experiments the study investigates the relative roles of abiotic and biotic processes in the development of tidal landscapes. Experiments are indeed novel as, for the first time (to my knowledge) they combine the analysis of the role of vegetation on the channel network formation and development in an intertidal basin context.

The experiments highlighted the role of vegetation in driving the development and evolution of channel networks investigating the differences associated with two different mechanisms of plant colonization (patchy and hydrochorous). Based on my personal experience I know the effort necessary to design, set up, and carry out physical experiments which made me further appreciate the study. I particularly appreciated the interesting way of concluding the introduction saying what the Authors expect (hypothesize) from their experiments.

Globally speaking I can say I recommend the manuscript for publication in Earth Surface Dynamics after minor revision. I provide the following few suggestions that are mostly issues of clarification and comments aimed, I hope, at further improving the readability and the quality of an already very good paper. No need to stay anonymous I am Luca Carniello.

SPECIFIC POINTS

Line 45: “Such channels are ebb-dominant in the unvegetated state, …” I suggest supporting this statement in some way. Being in the introduction the best option is to add some references.

Line 60: “history traits, such as plant recruitment strategies, can influence” I suggest adding two commas.

Line 160: “The first 10 tidal cycles were necessary to wet the tidal basin and re-establish a normal flow pattern.” Why is this necessary? Did you stop the experiment before each sowing event? I suppose not as at the beginning you state "During the experiment, Lotus seeds were dropped..." so, why do you need to re-establish the flow pattern? please clarify.

Line 165: “Around 160 000 seeds (i.e., 200 g) were supplied per sowing event to obtain a vegetation cover equaling about half the tidal basin at the end of the experiment.” Please explain why 160000 seeds are necessary to cover half of the basin.

Line 170: “while tilting of the flume was halted.” How long was the tilting halted? For the Hydrochloric sowing experiment, you say the experiment was stopped for 4 days. What happened when adopting the patchy sowing procedure? This important piece of information is actually missing.

Line 190: Can you explain the rationale that suggested you to run longer only one of the experiments with vegetation to check if the morphodynamic equilibrium was indeed reached? Why did you choose the patchy one? This is just a curiosity.

Line 201: “First, the raw laser line scanner data underwent a calibration and correction process for the laser-camera system.” Can you specify a little bit more in detail what the calibration and correction process consists of?

Line 201: “If the difference between the window median and the local pixel elevation was below a certain threshold (respectively, 0.0015, 0.005 and 0.0055 m) for at least one window size, the local pixel was identified as a channel.” I do not understand why this occurrence can ensure the selected pixel is in a channel. I suppose that the difference between the window median and the local pixel elevation can be below a certain threshold also for pixels pertaining to the adjacent flat areas. Can you clarify, please?

Line 286: “we observed that the left channel bend (as seen from vertical top-view)” I guess this is to explain to the reader what you mean by "left" channel bend but it is not clear to me what vertical means. I suggest defining left and right for example assuming an observer looking the the experiment from the inlet landward.

Line 337: I suggest remembering here that “DL” is the local drainage densities. It has been defined quite far above and I personally forgot.

Line 380: “In the hydrochorous seeding experiment, the vegetation cover increased slowly over time and remained lower than in the patchy seeding experiment.” It would be very interesting the investigate the effect of increasing the amount of seeds supplied per sowing event performing other hydrochorous seeding experiments. This is of course not a request of integration for this contribution but a suggestion for a further paper.

Review of ‘On the relative role of abiotic and biotic controls on channel network development: insights from scaled tidal flume experiments’ by Hautekiet et al., for Esurf

Overview: This study focuses on understanding biotic vs abiotic controls on tidal marsh development using four experiments. They conducted two unvegetated controls and two vegetated experiments, one with vegetation colonization by water and the other by air, with the goal of examining how these different experiments evolve channel networks. The authors step through a variety of metrics to examine the development of vegetation patterns and channel networks, concluding that channel network development is dominated by hydrodynamics near the sea and more strongly influenced by vegetation as you move landward. The paper is well-organized and easy to follow with nice figures. Below I outline a number of suggestions to improve the paper, which I believe constitutes minor revisions.

Comments:

The background is notably missing some references of previous vegetated experiments. Though tidal experiments with plants were not necessarily performed, there is a broad body of literature that informs the results found here, including experiments of deltas with vegetation (e.g., Piliouras et al., 2017, Piliouras and Kim 2018, 2019) and experiments at a smaller scale examining flow through and around patches of vegetation (e.g., Chen et al., 2012; Meire et al., 2014; Ortiz et al., 2013; Zong and Nepf 2010, 2011, 2012). These papers provide context to the results in discussing many of the processes at play, such as how flow is steered around patches of vegetation, how patches may encourage bifurcation to increase channel density, and how vegetation influences topographic change. I suggest including a paragraph in the background to elaborate on what we know about vegetation-flow-sediment feedbacks and drawing on this knowledge in the discussion, as well.

I would appreciate more information on the selection of this vegetation species, as it is not widely used in the literature for morphodynamics experiments, in addition to more information on the vegetation protocols. Can you provide some context on plant traits and impact on sediment transport? L165: You state 160,000 seeds to obtain coverage of half of an area – with what intended density? Is this the same density as in the other vegetated experiment? How was this density chosen? What impact might this have on your results? All of this should be justified in the manuscript. Finally, how was the timing of sowing determined? You’ve got 1000 cycles unvegetated, 10 cycles for wetting (is it not already wet?), 25 cycles to release, 25 cycles to spread, 500 cycles of just water/sediment transport, then more sowing – how were all of these things determined? You describe the experiments as scaled – are these numbers scaled to some timing of seed dispersal in a natural system? How was the number of patches and patch size determined for the manual sowing? Please address these issues in the manuscript text.

L193: How did you determine if equilibrium was reached?

Section 3.2: It seems from your methods that you are intentionally not sowing seeds in channels and not re-sowing if channels eroded through parts of patches. It’s worth a sentence or two in this section to explain that in the context of your results because you are pointing out things like channel bends having the lowest vegetation cover, but this is not surprising given your methods that intentionally do not sow in channels.

Figure 9 and other similar figures: It would help to differentiate the runs a bit more by changing line style and marker shape, in addition to color. This will likely also help folks who have difficulty discerning color.

I strongly suggest including movies of the experiments in supplemental. The results section is quite difficult to follow without having the movie available to the reader, especially when you talk about specific channels developing in specific runs and how that affected various metrics.

386-388: though they were not tidal experiments, this result is quite similar to that found by Piliouras et al., 2017, that rapid vegetation colonization can inhibit channel network development. Their later papers also found similar results about plant patches encouraging channelization and bifurcation around them. I suggest including these citations, as described in more detail above.

Figure 12: what are the gray shaded regions?

Regarding the phrasing of your conclusion that vegetation ‘dominates’ in the landward zone – does it really dominate? Or does it just have additional influence here? It seems that channel development is still a physical process that is initiated by abiotic factors, but that the initiation and evolution of the network is more strongly controlled by the combination of biotic and abiotic factors in the landward region, compared to a real dominance of abiotic factors near the coast. I suggest rewording this.

Data are not accessible. Please place data in a repository with DOI.

References:

Meire, D., Kondziolka, J. M., & Nepf, H. M. (2014). Interaction between neighboring vegetation patches: Impact on flow and deposition.

Water Resources Research, doi: 10.1002/2013WR015070

Ortiz, A. C., Ashton, A., & Nepf, H. (2013). Mean and turbulent velocity fields near rigid and flexible plants and the implications for

deposition. Journal of Geophysical Research: Earth Surface, doi: 10.1002/2013JF002858

Piliouras, A. and W. Kim. (2019). Upstream and downstream boundary conditions control the physical and biological development of river deltas. Geophysical Research Letters. doi: 10.1029/2019GL084045

Piliouras, A. and W. Kim. (2018). Delta size and plant patchiness as controls on channel network organization in experimental deltas. Earth Surface Processes and Landforms. doi: 10.1002/esp.4492.

Piliouras, A., W. Kim, and B. Carlson. (2017). Balancing aggradation and progradation on a vegetated delta: The importance of fluctuating discharge in depositional systems. Journal of Geophysical Research – Earth Surface. doi: 10.1002/2017JF004378.

Zong, L. J., & Nepf, H. (2010). Flow and deposition in and around a finite patch of vegetation. Geomorphology, doi: 10.1016/j.geomorph.2009.11.020

Zong, L. J., & Nepf, H. (2011). Spatial distribution of deposition within a patch of vegetation. Water Resources Research, doi: 10.1029/2010WR009516

Zong, L. J., & Nepf, H. (2012). Vortex development behind a finite porous obstruction in a channel. Journal of Fluid Mechanics, doi: 10.1017/jfm.2011.479

Dear editor,

Thank you for inviting me to review this interesting research that aims to understand the relative roles of biotic and abiotic controls on tidal channel network formation in a controlled, scaled, laboratory experimental setting. A total of four experiments were conducted at the state-of-the-art Metronome facility. These experiments included two unvegetated controls and two experiments featuring vegetation with different colonization strategies (patchy = random and hydrochorous = flow-driven). The resulting experimental tidal channel networks were analyzed using various metrics to examine the development of vegetation patterns and channel networks, leading to the conclusion that channel network development is dominated by hydrodynamics near the sea and is more strongly influenced by vegetation moving landward. The paper is well-organized and easy to follow, with excellent figures. The introduction section is particularly commendable for its clarity and conciseness. The results are novel and convincing for the most part (see comment below though), and overall of broad interest to the scientific community.

I can surely recommend publication, although I would first like the authors to make some clarifications and revisions. Below are my detailed comments:

1. Are the experiments reported here representative of tidal networks cutting through salt marshes or, more broadly, large tidal embayments? It seems they might be more representative of networks developed at the scale of entire tidal embayments (e.g., a back-barrier lagoon). The introduction supports the former, but upon reading the article and examining the figures, the latter seems more accurate.
2. Similar to experiments by Stefanon et al. (2010, 2012), the width of the tidal inlet is fixed a priori and kept constant throughout the entire experiment. This design has advantages, such as avoiding interference with flume walls and related boundary effects. However, it may prevent the system from fully adjusting morphodynamically throughout the experiment, potentially causing over-deepening at the inlet. Please discuss this point adequately.
3. Considering the above point, since waves were generated at the open sea boundary, why not allow barrier islands and tidal inlets to evolve naturally, adjusting morphodynamically during the experiments? Previous experiments on river deltas with waves and tides have shown that the dynamics of barrier islands and inlets can be replicated even at smaller scales than in the present experiments. See for example:
   
   Baumgardner, S.E. Quantifying Galloway: Fluvial, Tidal and Wave Influence on Experimental and Field Deltas [Ph.D. Thesis], University of Minnesota: 113 p, 2015. https://conservancy.umn.edu/handle/11299/183395
4. Waves are rarely discussed in the text and are not mentioned in the discussion at all. Given the fixed inlet, it's worth considering the purpose of including waves, especially since they aren't discussed. Please clarify.
5. The differences between control experiments and experiments with vegetation appear to be less pronounced than initially hypothesized and traditionally suggested in the literature. I'd like the author to comment on this. My impression (which might be entirely wrong) is that the differences are, in fact, so small that they could be due to the stochastic nature of the experimental results rather than a real genuine effect of vegetation. In other words, these differences might be noise rather than a signal, and averaging results over several repetitions of the same experimental run might amplify the results even further than they already are. (Please note that I AM NOT recommending running the experiment multiple times. I understand the effort required to conduct experiments at this scale, and reproducibility is challenging due to time and resource constraints).
6.  Line 395: "A slight reduction in tidal prism" raises questions about its feasibility. Since the width of the inlet is fixed, reducing tidal prism would require decreasing depth. I suspect that the inlet can hardly become shallower during the course of an experiment; in fact, it likely tends to deepen continuously until equilibrium depth is reached. Please provide clarification.
7. Line 420: "The more spatially homogeneous and hence weaker hydrodynamics in bare systems may be responsible for the lower degree of channelization." This holds true because we are examining the landward site of the basin, where hydrodynamics are weaker regardless of vegetation. Bare, unvegetated parts of a tidal basin typically experience higher hydrodynamic stresses (waves+tides). For instance, tidal networks develop on bare mudflats where hydrodynamics are dominated by sheet flow, with inertia playing a comparatively more significant role than in vegetated salt marshes.
8.  Line 445: "In conclusion, our results suggest a zonal domination of abiotic processes at the seaward side of intertidal basins with high hydrodynamic energy levels. In contrast, biotic processes dominate system development more toward the landward side with intermediate hydrodynamic energy levels." This conclusion holds for a large-scale tidal basin, such as the one reproduced by the model (see to comment n.1). However, I am curious whether these results can be downscaled to study tidal networks cutting through individual marsh islands, such as those shown in Figure 1, where energy levels are consistently low, and there is no tidal inlet, fixed or freely evolving (again, see reloated comment n.1)
9. Minor: The text consistently uses the number of tidal cycles as a proxy for time, which is generally acceptable, given that it is known that the tidal period T= 40 seconds. However, including references to the actual duration of the experiments at various points in the text would improve clarity, eliminating the need for readers to search for T value and calculate the actual time duration.
10. Minor: a relevant reference to recent experimental work on tidal channel network formation is missing:
    
    Geng, L., Gong, Z., Zhou, Z., Lanzoni, S., and D'Alpaos, A. (2020) "Assessing the relative contributions of the flood tide and the ebb tide to tidal channel network dynamics." Earth Surf. Process. Landforms, 45: 237–250.
11. Minor (but important): I support the comments from other reviewers suggesting the inclusion of movies of the experiments in the supplemental material to provide a better understanding of the experimental runs. Furthermore, making the experimental data freely accessible by placing them in a public repository with a DOI would be highly beneficial for the whole community.