General comments: 

Díaz-Astudillo et al use a 26 hours of acoustic data over a submarine canyon to provide another example of the importance of these topographical features in marine ecosystems. I commend the authors on their work. I think that is an interesting contribution and I have no specific comments on the validity of the methods or results. Overall, I think the manuscript is good, but I have two major comments that I would like the authors to address. 

Specific/Technical Comments: 

1) A majority of the text is verbose, making the text difficult to get through. I would encourage the authors to read the text aloud, or move any extraneous text to the supplement (while important, the extensive description of the wind/upwelling conditions and accompanying figures at the start of the results comes to mind) to consolidate the text to make it easier for the reader to understand your main points. I would also suggest that the authors reduce the number of figures and should take the time to thoroughly proofread the manuscript (ie make sure all abbreviations are consistently used throughout and verbs are in the correct tense). 

2) The authors put a lot of weight in their results, suggesting that they have discovered a potential fourth mechanism for submarine canyons facilitating food web focusing using just over a day's worth of data. I think both mechanisms 3 (interaction of DVM and abrupt topographies) and 4 (canyon mediated retention) are valid, but the fourth mechanism is not new. The Hudson et al 2022a, b papers (in MEPS [cited already] and JGR Oceans) illustrate both of these mechanisms. These mechanisms differ slightly than those described here (ie, the Antarctic krill are not being 'blocked' from migrating, or if they are, it was not considered as a focusing mechanism), but the overall concepts are the same. I would encourage the authors to combine these two mechanisms into a single section in which they describe how BBC retains zooplankton and then describe the ecological implications in a different section. 

Furthermore, there is a lot of weight put on these results (ie, suggesting a "new" mechanism) on 26 hours worth of data, which included only 2 zooplankton samples. We always want more data, and I still think that the study is strong with this limited dataset, but I think this needs to be addressed in the discussion and should include a brief (~1 paragraph) discussion of the limitations of this study. You performed this sampling study after a period of upwelling winds; discuss how you might expect this mechanism to change under different conditions. What if there hadn't been upwelling winds prior to your sampling event? If this mechanism is tied to upwelling winds, is it truly a different mechanism? 

This is an interesting study based on a very small dataset consisting of 16 transects with a shipboard 150khz ADCP and 4 zooplankton tows over the course of 26 hours. Although the dataset is limited in spatial and temporal extent the authors’ analysis provides some interesting insights into biophysical processes in a canyon within an upwelling regime.

I think the authors are describing a mechanism for zooplankton accumulation in the canyon not just retention? This concept with in the literature context of canyon funneling of zooplankton should be included in the scope of the paper.

It would be helpful to include a dashed line marking local sunrise and sunset in all figures that display time and all figures should be  bigger

Conceptually my biggest concern is the use of the term topographic blocking , which is actually quite specific and well defined in the long history. Topographic blocking specifically refers to vertical migrators whose sunrise descents are blocked by seafloor with depths shallower than their preferred daytime depths. The authors’ data does show some evidence for this at WT with what my best guess is for local sunrise (both sunrise and sunset times need to be explicitly stated in the text and marked in figures). Eg discussion ~l272. I think canyon funneling - where currents advert and accumulate animals at the canyon head should be explicitly discussed and relevant literature included new and old (eg greene et al 1988). This second type of topographic blocking is not topographic blocking but rather physical aggregation through advection toward the seafloor.

Overall time needs to be examined in more detail for the biological analysis and interpretation. Sunrise and sunset are important and will dramatically influence the density and distribution on zooplankton in the canyon. As is, the times of sunrise and sunset are not even explicitly mentioned.

Line specific comments

Figure 1 the color scheme for the bathymetry map is very confusing. There should not be two whites in one gradient. I suggest a single yellow to white gradient for land and a single gradient from white to dark blue for the seafloor. The cross canyon -rofilemis very useful . I would suggest adding the approximate locations of the zooplankton sampling stations.

L96 casts not transects?

L98 cite supply table 1

Section 2.4 Small fish with swim bladders will also scatter at 153khz - these results are not only for zooplankton.

L156 mark these 3 layers with dashed lines in all relevant plots for easy reference

Section 2.5 more details needed. Why not include a metadata ramble for the zooplankton sampling, location, depth, time, length of tow, volume sampled, biomass etc. were there 4 samples or 2?

How can you get 2 depth discrete samples? Did the Tucker trawl have 2 nets or 2 cod ends such that one was closed at a specific depth? How was sampling depth monitored? Were two separate casts done to sample the different depths? How did you avoid sampling on the way down and up? Opening and closing nets?

L179 the 1025 ‘isopycnal’ add isopycnal

Fig 5 caption would suggest defining positive cross shore direction and Spotify along shore direction in caption (onshore and equatorward)

Figure 6 p values and correlation values referenced but not shown in figure or written in text. These should be added

Suppl fig 5 is very useful but needs times, WT/ET labels, sunrise and sunset marked or identified in some way , NS/SS identified . It looks like in WT transects there is some evidence of sunrise accumulations of scatterers over the canyon walls. This would be evidence for topographic blocking as mentioned in the discussion. Because topographic blocking is specifically mentioned , these observations should be included in the text.

L226 which results? Also a 1m tucker trawl likely would not do very well catching fishes but that doesn’t mean that they aren’t contributing to the sound scattering .

It would be useful to combine figures 7 and 10 in order to be able to look at currents and mean sv side by side, especially given the authors arguments that the currents are structuring the zooplankton distribution.

Figure 8 add NS and SS as well as dashed lines for the 3 depth layers referenced. Consider

Figure 9 could be very useful with just a few changes : color code the times in a logical way - make all the nighttime transects blues/greens and the daytime transects reds/yellows then you will see that there is an interesting temporal pattern in these data! There are interesting changes in how topography influences sv over time - this may also strengthen the topographic blocking argument and should be added to the results in l232

L240 was the fact that sv are on a log scale taken into account before averaging ?

L245 I disagree with this Interpretation. SS of WT has a 20m thick layer during the day that greatly expands in thickness during the night

L284 seems odd to have references here -is this not the data presented in this manuscript? Mackas 2005 was not in BBC for example.

L285 don’t the data presented here show this? One thing that was missing is the  specific link over time between zooplankton layers and currents. What does your data tell you about how a migrating zooplankton in each of the 3 layers is adverted around the canyon over the 24 hrs? Where did a zooplankton who vertically migrated but was advected in the horizontal by the currents end up? Was there evidence for a funneling effect and a net transport into the inshore /canyon head direction? Were they advected in a circle and so were effectively retained ? How about a non migrating plankton at each depth. Making these points with given animal depths at different times and DVM behavior would make the biophysical connection explicit