Inadequate Sampling Frequency and Imprecise Taxonomic Identification Mask Results in Studies of Migratory Freshwater Fish Ichthyoplankton

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Dear Editor,

The authors correctly addressed the few issues still raised by Reviewer 3 and the Guest Editors. These corrections further improved the manuscript, which I believe can now be accepted on its present state.

Reviewer 2 Report (Previous Reviewer 3)

Dear editors, I checked the author's response and consider myself satisfied with his comments. In this way I believe that the work can be accepted for publication.Yours sincerely

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

GENERAL COMMENTS

I found this manuscript quite interesting and focused on an issue of increasing importance – the distribution of spatial and temporal variation of fish eggs assessed by DNA metabarcoding -  for fisheries science in the neotropics. The manuscript is overall well-written and structured with clear take-home messages. My principal concern is that many of the results, from example those concerning figures 3-6 should be supported with statistical tests, to check if there significant differences between the means. A simple ANOVA could do this, while adding (or not) support to the results. Table 1 should further include percentages of relative abundance (as it is referred in the text) and also further info on species ecology, e.g. the ecological guilds, as they are later referred on the Discussion. No doubt this would further improve the manuscript.

SPECIFIC COMMENTS

L4- remove the dot from title.

L83 – remove space in the word “Neotropical”.

L105 – Suggest adding “Fish” before “spawning”.

L107 – What was the basis of choosing these sites?

L111 – Iguatama? It does appear on the map.

L114-115 – What do the different colour of sampling sites represent? Increase font size on the scale of the smaller map.

L124 – You mean the location of spawning sites?

L206-207 – “48,465” Do these eggs included eggs other than of fish?

L214-216 – Table 1 showing species list should have a 2^nd column of percentages regarding relative abundance as you outline on the text.

L220-221 – “and among different sites within each site (Figure 2).” Please revise sentence. Among different reaches within each sampling site?

L223 – Have you tested the statistical significance of the means (for example with an ANOVA)? They seem very similar.

L248-251, 258 – What you mean by red/green/blue/gray simulation? This was not previously referred on the Material and methods.

L249 – “similar importance”. How was importance evaluated? Which indicator was used to evaluate importance?

Captions to figure 7 and 8 – What do the circles, lines, boxes,…represent?

L295 – undetected instead of un detected.

L295-300- Table 1, besides percentage abundance that should be included, could further include other columns outlining habitat, reproduction and parental care guilds.

L329- Hydroelectric?

 

 

Author Response

GENERAL COMMENTS

I found this manuscript quite interesting and focused on an issue of increasing importance – the distribution of spatial and temporal variation of fish eggs assessed by DNA metabarcoding -  for fisheries science in the neotropics. The manuscript is overall well-written and structured with clear take-home messages. My principal concern is that many of the results, from example those concerning figures 3-6 should be supported with statistical tests, to check if there significant differences between the means. A simple ANOVA could do this, while adding (or not) support to the results. Table 1 should further include percentages of relative abundance (as it is referred in the text) and also further info on species ecology, e.g. the ecological guilds, as they are later referred on the Discussion. No doubt this would further improve the manuscript.

Thank you for your constructive contributions. For each graph, statistical tests are now presented, and the table has been modified to include the suggested information. All the minor points have also been changed to accommodate your suggestions. Once again, we appreciate the suggestions that indeed enhance the manuscript.

 

SPECIFIC COMMENTS

L4- remove the dot from title.

Thank you. The suggestion was incorporated to the text. 

L83 – remove space in the word “Neotropical”.

Thank you. The suggestion was incorporated to the text. 

L105 – Suggest adding “Fish” before “spawning”. OK

Thank you. The suggestion was incorporated to the text. 

L107 – What was the basis of choosing these sites?

The distribution of the sites aimed to encompass all major tributaries of the basin, as well as intermediate stretches of the São Francisco River. Such information was included at the end of the paragraph.

L111 – Iguatama? It does appear on the map.

Thank you. We have changed the description to “upstream of the Bambuí river mouth”.

L114-115 – What do the different colour of sampling sites represent? Increase font size on the scale of the smaller map.

The colour of sampling sites were standardized and the font size on the scale of the smaller map increased.

L124 – You mean the location of spawning sites?

Thank you. We mean the location of spawning sites and the suggestion was incorporated to the text.

L206-207 – “48,465” Do these eggs included eggs other than of fish?

Thank you. The samples included only fish eggs and we included that on the text

L214-216 – Table 1 showing species list should have a 2^nd column of percentages regarding relative abundance as you outline on the text.

Thank you. Such an information was included in the table

L220-221 – “and among different sites within each site (Figure 2).” Please revise sentence. Among different reaches within each sampling site?

Thank you. We have changed to “among different samples within each site”.

L223 – Have you tested the statistical significance of the means (for example with an ANOVA)? They seem very similar.

Thank you. We have now tested the statistical significance of the means and they are presented (KW = 2.47; P = 0.01).

L248-251, 258 – What you mean by red/green/blue/gray simulation? This was not previously referred on the Material and methods.

We have removed the colours from the text description.

L249 – “similar importance”. How was importance evaluated? Which indicator was used to evaluate importance?

We have changed to “density of migratory fish eggs”.

Captions to figure 7 and 8 – What do the circles, lines, boxes,…represent?

As presented on the upper left of the figure, they represent the median, 25% and 75% percentile and range (min – max) respectively.

L295 – undetected instead of un detected.

Thank you. The suggestion was incorporated to the text.

L295-300- Table 1, besides percentage abundance that should be included, could further include other columns outlining habitat, reproduction and parental care guilds.

Thank you for the suggestion. We have included information on percentage abundance, migratory behaviour and care guilds.

 

L329- Hydroelectric?

Thank you. Yes Hydroelectric, the suggestion was incorporated to the text.

Reviewer 2 Report

In this study, authors explored the methods for analyzing the spawning situation of fish in river. That is valuable for environmental conservation. However, there are several vital problems in this study. First, the ichthyoplankton can not represent the whole spawning situation of migratory fish. As well known, there are sunken fish eggs as well as floating fish eggs in natural world. The ichthyoplankton may only represent the floating fish eggs, while the sunken fish eggs stay in the bottom of river. Secondly, samples were collected every 3 days from November 1st, 2019 to February 29th, 2020, and the conclusion that inadequate sampling frequency can mask results of the study in the ichthyoplankton was made just by the comparison of the results of every-3-day sampling, every-6-day sampling, and every-15-day sampling. However, the hatching periods of eggs from different species in fish are significantly various, which may range from 1 day to several days. High-frequency sampling may overlap the eggs of long-term hatching period species. It is not accurate to make a conclusion that every-3-day sampling is correct and every-6-day sampling or every-15-day sampling is masked. The DNA metabarcoding method was utilized to identify the spawning species in fish. That absolutely has advantage against the non-molecular identification. But the DNA metabarcoding method was conducted on the selected eggs. It is possible that there might be some migratory fish eggs which were ignored by artificial selection. The samples after artificial selection can not represent the actually spawning situation of migratory fish. What is more, the English language in this manuscript could be improved. This reviewer believed that this manuscript should not be published unless a fundamental revision on study methods was actualized.

it is ok。

Author Response

In this study, authors explored the methods for analyzing the spawning situation of fish in river. That is valuable for environmental conservation. However, there are several vital problems in this study.

I appreciate the criticisms of the manuscript, but I believe that most of them stem from a lack of knowledge about the migratory fish fauna in the studied system.

 

First, the ichthyoplankton can not represent the whole spawning situation of migratory fish. As well known, there are sunken fish eggs as well as floating fish eggs in natural world. The ichthyoplankton may only represent the floating fish eggs, while the sunken fish eggs stay in the bottom of river.

With the exception of the Amazon Basin, most South American rivers are shallow. In these systems, which include the basins of the São Francisco and the Plata rivers, for example, the total migratory species have free-floating, near-buoyant eggs that are distributed homogeneously in the water column. Since floodplains serve as the primary nursery for these species, the eggs closest to the water surface are more likely to be transported to these flooded regions. For this reason, surface ichthyoplankton sampling is the one used in these basins' surveys (see references below). Due to these reasons, we are convinced that the sampling design was the most appropriate for the upper São Francisco.

MELO-SILVA, MERIELI ; DA SILVA, JISLAINE CRISTINA ; Bialetzki, Andréa . Community structure of fish larvae in different biotopes of a neotropical river. COMMUNITY ECOLOGY, v. 22, p. 1, 2022.

BRAMBILLA, EDUARDO MENEGUZZI ; SILVA, LUIZ G. M. ; BAUMGARTNER, LEE J. ; Bialetzki, Andréa ; NOGUEIRA, Marcos Gomes . Dispersal of fish eggs and larvae in a cascade of small hydropower plants with fish ladders. HYDROBIOLOGIA, v. 1, p. 1, 2020.

ROSA, G. R. ; SALVADOR, G. N. ; BIALETZKI, A. ; SANTOS, G. B. . Spatial and temporal distribution of ichthyoplankton during an unusual period of low flow in a tributary of the São Francisco River, Brazil. RIVER RESEARCH AND APPLICATIONS, v. xx, p. 01, 2017.

SOARES, M. L. ; MASSARO, MARTHONI VINICIUS ; HARTMANN, P. B ; SIVERIS, S. E. ; PELICICE, F. M ; Reynalte-Tataje, David A. . The main channel and river confluences as spawning sites for migratory fishes in the middle Uruguay River. Neotropical Ichthyology, v. 20, p. 1, 2022.

SILVA, FERNANDA DE OLIVEIRA ; ANDRADE NETO, FRANCISCO RICARDO DE ; HERMES SILVA, SAMARA ; SILVA, JADE DE OLIVEIRA DA ; Zaniboni Filho, Evoy ; PRADO, IVO GAVIÃO ; PERESSIN, ALEXANDRE ; PELICICE, FERNANDO MAYER . Recruitment dynamics of a migratory fish in a semi-arid river system. Inland Waters, v. 10, p. 1-13, 2020.

 

Secondly, samples were collected every 3 days from November 1st, 2019 to February 29th, 2020, and the conclusion that inadequate sampling frequency can mask results of the study in the ichthyoplankton was made just by the comparison of the results of every-3-day sampling, every-6-day sampling, and every-15-day sampling. However, the hatching periods of eggs from different species in fish are significantly various, which may range from 1 day to several days. High-frequency sampling may overlap the eggs of long-term hatching period species. It is not accurate to make a conclusion that every-3-day sampling is correct and every-6-day sampling or every-15-day sampling is masked. 

Once again, the raised concern does not apply to the studied system. The hatching of South American migratory fish eggs occurs extremely rapidly. For all the species studied, the hatching time is less than 24 hours. To dispel any doubts among the readers, we have added the following sentence: "Since the hatching time for migratory fish in the São Francisco basin is always less than 24 hours (Sato et al., 2003), the 3-day collection interval ensures that each sample represents an independent spawning event.”

 

The DNA metabarcoding method was utilized to identify the spawning species in fish. That absolutely has advantage against the non-molecular identification. But the DNA metabarcoding method was conducted on the selected eggs. It is possible that there might be some migratory fish eggs which were ignored by artificial selection. The samples after artificial selection can not represent the actually spawning situation of migratory fish.

There were no artificial selection, since all the eggs were analyzed by DNA metabarcoding. As described in lines 130-132 “All samples containing fish eggs were analyzed using the DNA metabarcoding method, which allows the analysis of various organisms combined in a single sample through high-performance DNA sequencing (Illumina).”

 

What is more, the English language in this manuscript could be improved. This reviewer believed that this manuscript should not be published unless a fundamental revision on study methods was actualized.

 

The manuscript was reviewed by a native speaker

Reviewer 3 Report

The article: “Inadequate sampling frequency and imprecise taxonomic identification can mask results in studies of migratory freshwater fish ichthyoplankton” clearly shows some assumptions that for those who work with the ecology of fish from continental waters are almost obvious: identification through DNA is a fundamental tool for identification and that a greater number of days of ichthyoplankton collection will always be better than one or a few days of collections in the month, mainly when one wants to detect eggs of migratory fish that have total spawning. The clear ideal would be to collect every day of the month and the entire reproductive period. Despite this, the work is a good reminder and a warning that the methodologies currently used, mainly in monitoring studies and ichthyofaunistic surveys, could be improved, both in terms of tools (inclusion of DNA analysis) and in the periodicity of collections.

The work, although very well written and discussed, has a serious flaw: using the assumption that the nets only capture semi-dense eggs and that they are all migratory fish eggs. The criterion of semi-dense egg with high perivitelline space size (>30% of total egg diameter (DE)) had to be used to identify migratory fish eggs. The most used identification manual in Brazil: Nakatani et al. (2001) and other authors, including some cited in this article (see Soares et al., 2022 and Becker et al., 2015) use this criterion. The work by Becker et al. (2015), also carried out in the São Francisco river basin and who also compares the efficiency of using DNA methods with respect to traditional egg identification methods, accurately considers this criterion when comparing the two methods. Certainly many of the eggs captured in this work would have a perivitelline space less than 30% of the DE.

Therefore, from my point of view, considering all eggs captured by the net was an erroneous criterion. So much so that, for example, Anchoviella eggs that are known to have oil droplets and are not semi-dense (they are pelagic) were also considered. Thus, the result obtained in the article disproportionately shows the efficiency of the method using DNA to the detriment of the traditional method.

Round 2

Reviewer 1 Report

Dear authors,

Thank you for properly addressing my comments on this revised version. I am overall satisfied with the changes made, and the manuscript can now be accepted as it stands. Well done!

Reviewer 2 Report

The answers are still lacking reliable materials. So, the present paper is not suitable for accept in my opinion.

it is ok.