Target Strength Measurements of Free-Swimming Sandeel Species, Ammodytes spp., in a Large Indoor Experimental Aquarium

Round 1

Reviewer 1 Report

Please see the attachments.

Comments for author File: Comments.pdf

Author Response

Thank you for your valuable comments. We have fixed all of your points and added explanations.

1. Page 1, 35~36 lines, it would be good to write down the scientific names of the three species of sandeels that are actually distributed along the coast of Japan.

 

   We have added the scientific names of the three species sandeels to the text.

2. Page 1, 39~42 lines, I think that the sandeels (Ammodytes spp.) are an important fish species all over the world, more emphasis on this importance in the paragraph will highlight the TS measurements.

 

We have changed the text to [In all sea areas, the sandeel occupies a critical ecological position in the conversion of lower trophic levels into higher ones, because it feeds on zooplankton and is a major prey of marine mammals. Therefore, sandeel is an important mid-trophic species in the sea ecosystem. In Japan, sandeel is widely distributed along the coastal area and is a vital target resource for coastal fisheries nationwide].

And we also added the new paper which researched in UK (https://doi.org/10.3354/meps337279).

 

3. Page 4, 140~152 lines, why are the author estimating mean and maximum TS values and relationships with FL?

 

Regarding the fork length FL, we added a note in the text as follows: [Since fish in natural environment may experience caudal fin damage, the body length of TS equation used for future abundance estimates was the fork length, which is unaffected by caudal fin damage].

 

4. Page 4, 153~167 lines, the swimming angle is calculated by the split-beam of the 3-D coordinate. Why the underwater camera is not used?

 

In our study, we used swimming angle just to verify whether the TS_mean estimated in the experimental aquarium was close to the natural sea area. Because it is not necessary to use the title angle distribution for estimated TS_mean in our study, we do not used video cameras or other equipment to measure the title angle. We have added this information in the ‘Methods’ and discussed the difference between swimming angle and tilt angle in the ‘Discussion’. Additionally, we turn off the lights at night to create the same light/dark environment as outdoors. This makes it difficult to measure with a camera.

 

5. Page 7, 272~273 lines, why did the author compare only previous studies of A. personatus species? In this study, Ammodytes marinus species included in the TS measurement?

 

Previously, the A. personatus and A. hexapterus have been recognized in the waters surrounding Japan. However, a taxonomic reexamination based on genetic and morphological analyses has led to the description of a new species A. heian, and scientific name A. japonicus should be applied to the A. personatus. Based on the above, A. personatus is most likely the same species or closely related to the present study. In addition, A. marinus is not included in the TS measurements of this study because it is not found around Japan. We added the species information in the table 4 text.

 

6. How about the TS values compared this study with the previous study at 200 kHz (Kubilius and Ona, 2012, 1106 page of Figure 6)?

 

   Because TS are frequency dependent, even though the fish of the same length will have different values when measured at different frequencies. In addition, the higher the frequency, the greater the variation in TS. For these reasons, since the frequencies used in this study was 38 and 120, we compared only with previous studies at the same frequencies.

 

7. Page 3, Table 2, 104-105 lines, what is the reason for the net empty on the right side?

 

   We have narrowed the experimental range so that more data can be measured. We believe that the even current experimental range is sufficient for fish to swim freely.

 

8. Page 8 with Figure 5, 275~278 lines, why did a TS values difference between -0.9 and -10.2 dB at 38 kHz, -3.3 and -8.4 dB at 120 kHz occur. It needs enough explanation.

 

   We have written a more detailed explanation of this as follows [The theoretically estimated TS_mean was −64.9 ~ −63.3 dB at 38 kHz and −69.0 ~ −64.8 dB at 120 kHz, while the measured TS_mean was −64.8 ~ −54.2 dB at 38 kHz and −74.4 ~ −57.6 dB at 120 kHz. The measured values were higher than the theoretical values at both frequencies, and the difference between theoretical and measured value is larger at 120 kHz than at 38 kHz. One possible reason for the discrepancy between the theoretically estimated TS_mean and the measured TS_mean is that the parameters used in the DWBA model, such as tilt angle, may have differed from those used in the actual measurements. And the directionality becomes stronger at high frequency, even small changes in tilt angle cause large fluctuations in TS values. The tilt angle in the DWBA model was assuming a normal distribution with a mean of -5° and a standard deviation of 15°; however, because this study was free swimming, the tilt angle distribution may not be normal distribution. Therefore, the difference between theoretical and measured values may reflect the difference in tilt angle distribution, and large difference at 120 kHz may reflect the strong directivity at high frequencies].

 

9. The author said that the theoretical model input parameter, which is one of the causes of the difference, is not a big difference in the TS value for a small size, and why is there a big difference in the large size? Please add a more explanation for that.

 

Under normal circumstances, for all length classes, the TS values for the sandeel measured in this study should be as strong as the model estimates. However, it can be assumed that sandeel in the smaller size classes have finer swimming behavior than those in the larger size classes, resulting in a larger variation in TS for the smaller size classes and a smaller average TS. Therefore, it may be better to consider that the difference in TS between the two studies was not smaller for the small size glass and larger for the large size class, but rather that the measured values for the small size class were weaker and closer to the model measurements.

 

10. Page 8 with Figure 5, 280~287 lines, only the density (g) value was mentioned, is there no effect on the sound speed ratio (h)?

 

We have added the information about sound speed ratio (h) as follows [In addition, possible reasons other than tilt angle distribution are the density ratio g between the fish body and seawater and sound speed ratio h, which are the important parameters for TS estimation by the theoretical model. The DWBA model is used under the assumption that g does not change with fish body tilt; However, because actual fish bodies are composed of flesh and bones, g changes due to the change in tilt angle. Similarly for h, it can assume that the values of h used in this study and in the model are different because the same seawater is not used. Therefore, it is considered that changes in g and h have a large effect on difference of TS between model measurement and this study].

 

11. Page 6, Figure 3, 204 lines, please divide figure into a, b, c, and d.

 

We have divided the figure into 4 parts and expressed the TS-FL relationship with the figure.

 

12. Page 6, Figure 4, 212~214 lines, it seems easier to understand if the x-axis is displayed from -30 to +30 degrees and the histogram of bin size is presented at intervals of about 3 degrees.

 

   We have displayed the x-axis from -30 to +30 degrees and at intervals of about 3 degrees.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Review of:  Target strength measurements of free-swimming sandeel species, Ammodytes spp., in a large indoor experimental aquarium.

Date 040622

General comments:

The paper is reporting about the mean and maximum target strength of sandeel over several size groups, from measurements made on free swimming individuals in a large indoor tank. The measurements are made at two operating frequencies, 38 and 120 kHz. The paper is well written, and the experimental procedure seems sound and accurate, with interesting results. The paper has 12 ! authors, which is good if they can state their responsibility in the work.

The most interesting result is that the target strength to size relationship does not follow a general 20logL slope, but larger, if linear fitting procedures are used, also for the parameter (a).

The reason for making target strength measurements is clearly given in the introduction, as the mean TS is going to be used to convert acoustic data (Sv) to density in acoustic surveys to assess abundance and catch rates of sandeel stocks.

Some references to international literature is lacking, where, in some regions, acoustic survey assessments have been used with success for more than 10 years now.

When the mean target strength will be used in real assessment, it is important that the target strength measurements does not introduce large errors, since an error here will directly affect the biomass estimate.

Having earlier stated that I like the setup and analysis, somethings must be wrong, since their results end up with a mean TS for sandeel which is nearly 10 dB stronger than the earlier reported values, and subsequently, the biomass will be 10 times lower if adopted in the assessment.

The other observation is that the mean TS is stronger at 38 kHz than at 120 kHz, which is not expected for samall and intermediate sizes of sandeel, at least based on measured frequency response.

 

Specific comments:

 

Line 48 : …..when water temperature is high. (I believe that not only because of temperature, but also the availability of suitable zooplankton, like copepods).

Line 69:   … is required. ( to make good and valid in situ TS measurements on sandeel is very difficult, mainly due to its schooling density. Previous trials have been many animals together in cages (British 1980-ties), and some in situ in Bottom cages (Kubilius 2015) and the referred Japanese experimental measurements.  So, search the international literature, and summarize what is found on cages, ex situ, in situ and modelling works in the introduction.

Line 74: I think Kubilius (2015) made measurements on free swimming fish, but on bottom in a large cage, trapping sandeel as they came out of the sand.

Line 112, also state that some data came from 2021, mixed with data from 2017.

Line 112: Calibration is important and describe more in detail how this system is calibrated. Usually, for a split beam system, the sphere is guided through the acoustic beam, and the parameters for estimated center gain, G0, split beam opening angles, and beam offsets are estimated in the calibration procedure. If not part of the interlan system, this can easily be done in postprocessing.

Since we are measuring on weak single targets, I would be interested in your signal to noise ratio, SNR (dB), which should be 10 to 20 dB above your minimum TS or so, if an accurate representation of the TS distribution is expected.

Line 117: Was no raw data recorded, and only accepted TS data from the system itself recoded. If this is the case, no postprocessing or reprocessing is available ??

 

Line 138 to 148

What is the single echo detector settings, SED

How is mean TS calculated from the distribution?

How large part of the acoustic beam was used for TS measurements?

If the signal to noise ratio is not super, >30 dB, than sorting the accepted echoes with respect to THETA, in 0.5 or 1 degree sections may be used to estimate a beam cuttof for valid TS measurements.

Plotting TS as a function of THETA can also reveal problems with SNR in the outskirts of the beam.

When working with weak targets, as here, I think this is a wise procedure. See example in my capelin in situ TS data. (added PPT example).

Line 161Is swimming angle the same as tilt angle ??

You have been using Target tracking to determine the swimming angle. But for negatively buoyant fish, like sandeel, the progress through the water = swimming angle from tracking, may be linear horizontal, even with a high, 20 degree tilt. The only way to determine this is through video of the swimming behaviour (not mentioned here), and then determine of the root of the tail passes the same point as the snout, like done in Ona (2000) for herring. Of course, if swimming speeds are high, then the hydrodynamics of the animal will force this to be true, but at lower speeds, many fish can have very different tilt angle during straight swimming.

In acoustic backscattering, it is the tilt angle which is important, due to the directivity of the body.

 

Now I come to may main point of my investigation of the manuscript.

If we use our earlier knowledge on where to expect the target strength of sandeel, we know it is without a bladder, with similar backscattering from fish flesh as Atlantic mackerel. The mackerel TS has been investigated by many methods, and ahs a suggested mean TS = 20logL -86 dB re 1 m²

A 15 cm mackerel is considerably thicker and has higher weight than a 15 cm sandeel, so we should expect maybe half of the echo, or -3 dB lower for sandeel, or maybe a b₂₀ of -90 dB, or so. (If their orientation in the sea is the same). We should also expect a similar frequency response for small mackerel as for a intermediate sandeel. This is also confirmed by modelling.

So, taking this as a starting point, we should expect a mean TS for a 15 cm sandeel to be about -66.5 dB, but during your TS measurements, you used -70 dB as a lower threshold in the SED. And a -45 dB as a maximum. This means that you have removed most of the weak echoes from your analysis, and your mean TS is too high on both frequencies.

This is easily adjusted if your data is fully recorded, but lost if only the TS data is recorded !

I would recommend at least 10, but usually 15 dB, lower that the mean as the TS Minimum setting in the SED, and also with the check of how efficient you can record -85 dB in the outskirts of the beam, as described earlier.

If this is possible, you have very interesting data on TS size dependency for sandeel, which is worth publishing.

 

Conclusion:

Publishable if it is possible to reanalyse the data with lower TS minimum, and better description of the calibration procedure.

Also discuss better difference between swimming angle and tilt angle.

 

 

 

 

 

 

 

 

 

 

 

Comments for author File: Comments.pdf

Author Response

We really appreciate your valuable advice. We have lowered the threshold and reanalyzed the data. More information was also added on calibration and swimming angle.

1. Line 69:  Search the international literature, and summarize what is found on cages, ex situ, in situ and modelling works in the introduction.

 

We have changed the text to [Three methods are commonly used for TS estimation, including ex situ methods and modeling methods for live fish in cages and immobile fish, and in situ method for wild fish. The ex situ methods and modeling methods control the posture of the experimental species and cannot easily obtain TS data close to those obtained in the natural state. The in situ method can obtain TS in the natural state; however, the inclusion of other fish species introduces large TS errors because TS often cannot directly reveal the target fish measured. Generally, the TS values of fish are influenced considerably by having or not having swim bladder, and especially swim bladderless fish are more affected by body length and swimming behavior. Thus, to make good and valid TS measurements on sandeel, which is a swim bladderless fish, method of measuring TS while allowing the sandeel to swim near the natural conditions is required. However, it is extremely difficult to measurement swimming sandeel because its schooling behaviour and weak acoustic target. Previous studies of swimming sandeel were measured multiple individuals in cages in a tank or in cages under natural environment. Since the TS measured in a fixed space, it was difficult to detect individual target and the fish behavior is restricted. Therefore, to accurately estimate the abundance of sandeel, it is necessary to estimate TS in conditions similar to their natural swimming conditions, while considering various swimming behaviours. According to these, a method of measuring TS while allowing the sandeel to swim freely in a relatively large, physically controlled environment is expected.]

 

2. Line 112: Calibration is important and describe more in detail how this system is calibrated.

 

We wrote more about calibration as follows [Before the start of each experiment, the echo sounder was calibrated using a tungsten-carbide sphere (φ38.1 mm), which has a standard acoustic reflection with a known TS. The calibration sphere was hanged about 2-3 m below the transducer, and the transmission / reception coefficient (TR: value obtained by multiplying the voltage sensitivity by the gain) was calibrated comparing the measured TS to the theoretical TS. The seawater sound velocity and absorption attenuation used in the TS calculation were obtained by measuring the water temperature salinity values during the experiment].

 

3. Line 117: Was no raw data recorded, and only accepted TS data from the system itself recoded.

 

The raw data also was recorded and saved. We have changed the text to [Each experimental was recorded for more than 12 hours, and acoustic data, when fish pass within the beam, was recorded. The TS data identified by the single-target detection algorithm in the echo sounder were detected from the recorded acoustic data, and various information such as voltage phase difference of single echo in front-back and left-right directions of transducer, etc. are output with the TS data as CSV file].

 

4. How large part of the acoustic beam was used for TS measurements? Since we are measuring on weak single targets, I would be interested in your signal to noise ratio, SNR (dB), which should be 10 to 20 dB above your minimum TS or so, if an accurate representation of the TS distribution is expected.

 

 The single-target detection algorithm in our quantitative echo sounder uses a method that detects only those echoes that can ensure a signal-to-noise ratio (SNR) of 20 dB. The output results as CSV file are recorded only for single echoes recorded within a cutoff angle of 5.0° considering a -3 dB beam width. We have put this information in the text.

 

5. If the signal to noise ratio is not super, >30 dB, than sorting the accepted echoes with respect to THETA, in 0.5 or 1 degree sections may be used to estimate a beam cuttof for valid TS measurements.

 

    We checked the extracted TS data, and all data were contained within a cutoff of 0.5 degree. We calculated the beam degree from voltage phase difference of single echo in X axis (front-back direction) and Y axis (left-right directions) of transducer.

 

6. What is the single echo detector settings, SED. we should expect a mean TS for a 15 cm sandeel to be about -66.5 dB, but during your TS measurements, you used -70 dB as a lower threshold in the SED. And a -45 dB as a maximum. This means that you have removed most of the weak echoes from your analysis, and your mean TS is too high on both frequencies. I would recommend at least 10, but usually 15 dB, lower that the mean as the TS Minimum setting in the SED.

 

We have performed reanalysis. We determined the maximum TS value of background noise from the raw data of each experiment, and the maximum value of background noise in the range of -83 to -73 dB at 38 kHz and in the range of -87 to -84 dB at 120 kHz. Because the SNR of 20 dB is ensured by single-target detection algorithm, we set the lower threshold of the analysis to the maximum background noise for each experiment. We also put this information in the text.

 

7. How is mean TS calculated from the distribution?

 

The procedure for calculating the TS_mean was to convert the TS of each ping to a linear value, average them, and then reconvert them to dB. We have put this information in the text.

 

8. Line 161Is swimming angle the same as tilt angle?? You have been using Target tracking to determine the swimming angle. But for negatively buoyant fish, like sandeel, the progress through the water = swimming angle from tracking, may be linear horizontal, even with a high, 20 degree tilt. The only way to determine this is through video of the swimming behaviour (not mentioned here), and then determine of the root of the tail passes the same point as the snout, like done in Ona (2000) for herring. Of course, if swimming speeds are high, then the hydrodynamics of the animal will force this to be true, but at lower speeds, many fish can have very different tilt angle during straight swimming.

 

In our study, we used swimming angle just to verify whether the TS_mean estimated in the experimental aquarium was close to the natural sea area. Because it is not necessary to use the title angle distribution for estimated TS_mean in our study, we do not used video cameras or other equipment to measure the title angle. We have added this information in the ‘Methods’ and discussed the difference between swimming angle and tilt angle in the ‘Discussion’.

 

9. A 15 cm mackerel is considerably thicker and has higher weight than a 15 cm sandeel, so we should expect maybe half of the echo, or -3 dB lower for sandeel, or maybe a b20 of -90 dB, or so. (If their orientation in the sea is the same). We should also expect a similar frequency response for small mackerel as for a intermediate sandeel. This is also confirmed by modelling.

 

The results of the reanalysis are as follows:

the relationships between FL and TS_mean, and FL and TS_max at the two frequencies were obtained. With respect to the relationship between FL and TS_mean, in the case of no fixed coefficients, a and b were 50.2 and –120.4 dB (r² = 0.64), respectively, at 38 kHz, and 71.3 and –153.2 dB (r² = 0.62), respectively, at 120 kHz. When coefficient a was fixed at 20, the TS_cm values were –82.8 dB (r² = 0.40) and –89.2 dB (r² = 0.30) at 38 kHz and 120 kHz, respectively.

 

10. The other observation is that the mean TS is stronger at 38 kHz than at 120 kHz, which is not expected for small and intermediate sizes of sandeel, at least based on measured frequency response.

 

The same results are obtained even when the analysis threshold is lowered. The same result was also obtained for Yasuma et al. 2009 (The values of TS_mean were higher at 120 kHz than at 38 kHz in all juvenile fish examined (<6.7 cm), and values of ∆TS were over 5 dB. On the other hand, the values of TS_mean were higher at 38 kHz than at 120 kHz in adult fish (>7.5 cm), and values of ∆TS were from 0 to -6 dB). Because the experimental fish in this study are larger than 7 cm in body length, we think it is possible to obtain the same results

 

 

11. Also discuss better difference between swimming angle and tilt angle.

 

We have added the text about swimming angle and tilt angle as follows [The swimming angle can be viewed as an indicator of postural changes that affect TS; however, it is different from rigorous attitude angle. The orientation of fish changes in response to the direction in which they are moving, but at lower speeds, many fish can have very different tilt angle even though during straight swimming. Therefore, it is de-sirable in the future to conduct TS measurements under conditions close to natural seawater to verify the difference between the tilt angle and swimming angle].

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

According to the review, the answers and corrections seem to have gone well.

However, why are there no answers to all comments?

Comments for author File: Comments.pdf

Author Response

Minor comments:

1. Page 2, 62 line, three methods are commonly used for TS estimation.

 

We have changed the text to [Three methods are commonly used for TS estimation, including ex situ methods and modeling methods for live fish in cages and immobile fish, and in situ method for wild fish].

 

2. Page 3, 111 line, why does the author have a total of nine experiments?

 

Since the experiment date is 10 times, we have changed the text to [The fork lengths (FLs) of the specimens ranged from 13 to 22 cm and a total of ten experiments were conducted from 29 June to 10 July 2017, and from 15 to 23 June 2021 (Table 1)].

 

3. Page 3, 113 line, add a reference for standard acoustic calibration.

 

We have added the information and references for standard acoustic calibration as follows [Before the start of each experiment, the echo sounder was calibrated using a tungstencarbide sphere (φ38.1 mm), which has a standard acoustic reflection with a known TS (Sawada, 2002). The calibration sphere was hanged about 2-3 m below the transducer, and the transmission / reception coefficient (K_TR: value obtained by multiplying the voltage sensitivity by the gain) was calibrated comparing the measured TS to the theoretical TS. The seawater sound velocity and absorption attenuation used in the TS calculation were obtained by measuring the water temperature and salinity values during the experiment (Francios et al., 1982).].

 

4. Page 6, Figures 3 and Page 8, figure 5, expressing the x scale in log seems easier to compare.

 

The reason for using body length as the horizontal axis is so that the change in TS with body length can be visually determined. We think it is good to be able to determine at the moment of seeing which body length is changing abruptly. For this reason, we have chosen the horizontal axis to be the body length, not the log of body length.

 

 

 

Main comments:

5. Page 8, 296~297 lines, when estimating the TS equation, is it reasonable to compare adult and juvenile sandeels?

 

　　 Prior research has suggested that a possible method for estimating TS in juvenile fish is to apply the equation relating TS and body length obtained from measurements of adult fish to juvenile fish. However, an exception is made for fish with swim bladders. That is the relationship between body length and swim bladder volume (which reflects most of the TS) varies with growth stage. When swim bladder fish are small, swim bladder growth is seen to be faster than fish body growth.

Since the fish in this study do not have a swim bladder, it can be assumed that TS will be length dependent. Therefore, we believe it is worthwhile to compare TS equations of juvenile and adult fish.

 

6. Page 8, 298~302 lines, why was the comparison of a and b not with the mean and maximum TS values in the previous study?

 

The reason for using only TS_mean and comparing it to previous studies is the parameter typically used for resource estimation is TS_mean to avoid overestimating resources, not TS_max. Additionally, TS_max is not much affected by parameters such as directionality and swimming angle, so that it does not reflect significant differences from previous studies due to parameters. Therefore, we only focused on the a, b of TS_mean when compared this study to previous studies.

 

7. Page 8~9, Table 4 and Figure 5, 312~316 lines, it is necessary to accurately indicate whether the body length is the total length or the FL.

 

We have added the information of body length in the text, table and figure.

Author Response File: Author Response.pdf

Reviewer 2 Report

EDITOR WILL DECIDE

 

Author Response

We sent out English proofreading and cleaned up the text.

Author Response File: Author Response.pdf