Morphometrics, Growth and Condition of the Invasive Bivalve Rangia cuneata during Colonisation of the Oder Estuary (North-Western Poland)
Round 1
Reviewer 1 Report
The article deals with the important issue of invasive species in the environment. Against the background of global warming, their expansion can significantly disrupt the functioning of aquatic ecosystems. I recommend the article for publication after the following comments.
Line 26: "The Baltic Sea was formed about 20,000 years ago [1]." This sentence is false. There was an ice sheet in this part of Europe during this period. This should be changed and the source should be verified.
Please change the caption under the figures and the corresponding references. Applies to the entire text.
Line 80: Figure 1. Location of Rangia cuneata sampling sites in the Oder Estuary.
Line 131: Figure 1. Length structure of R. cuneata population in the Oder estuary.
etc.
The article lacks a broader environmental interpretation of the recorded size variation. What is the size of seawater intrusion and its spatial extent in the Oder estuary? What effect does this have on the obtained indices?
The authors point to the special role of water temperature, but there is no compilation of such data during the study period. The analysis of this parameter is too general in nature (Line 271-274).
Sentence : It 272 should be mentioned that water temperatures in the Szczecin Lagoon range from 0.5°C 273 (in winter) to 26°C (in summer), with an annual average of approximately 11oC [56]. The article was cited as the source of the information: Verween, A.; Kerckhof, F.; Vincx, M.; Degraer, S. First European record of the invasive brackish water clam 512 Rangia cuneata (G.B. Sowerby I, 1831) (Mollusca: Bivalvia). Aquat Invasions 2006, 1, 198-203. 513 DOI:10.3391/ai.2006.1.4.1. The studies collated there concern the Harbour of Antwerp (Belgium), and hence cannot be referred to the Oder estuary. This should be changed.
Also, what is the role of the temperature of the Oder estuary (southern sites) on the results obtained compared to the northern sites?
Line 323-324. "This section is not mandatory but can be added to the manuscript if the discussion is unusually long or complex". This is an editorial annotation. It should be removed from the text.
The conclusions section should address the question of further research. What research can or must be done?
References:
The references record is inconsistent with MDPI standards.
Author Response
Reviewer 1
Thank you very much for your review. We have made every effort to introduce all the corrections that were detailed in the review
Line 26: "The Baltic Sea was formed about 20,000 years ago [1]." This sentence is false. There was an ice sheet in this part of Europe during this period. This should be changed and the source should be verified.
Please change the caption under the figures and the corresponding references. Applies to the entire text.
We agree with the reviewer's suggestion. We checked the literature. Szaniawska (2018) writes that the Baltic Sea in the form known today was formed several thousand years ago, after the end of the last ice age:
We introduced: The Baltic Sea was formed about a few thousand years ago since the end of the last ice age) [1]...
Line 80: Figure 1. Location of Rangia cuneata sampling sites in the Oder Estuary.
We agree that the citation of Figure 1 in the text is misleading. We moved the quote to a place that better illustrates what we wanted to present.
We introduced: Between 12 November and 20 December 2022, 504 specimens of Rangia cuneata were collected during sampling of benthic organisms in the Szczecin Lagoon (Oder estuary) (Figure 1). using a Van Veen grab sampler within a sample area of 0.1 m2.
Line 131: Figure 1. Length structure of R. cuneata population in the Oder estuary.. etc
We agree that we have introduced incorrect numbering of figures and tables. We have corrected the numbering.
We introduced:
Line 131 Figure 4. Length structure of R. cuneata population in the Odra estuary
Line 172 Figure 6. Age structure of R. cuneata.
Line 196 Figure 9. Growth in length (A), height (B) and width (C) of R. cuneata shells together with growth parameters according to the von Bertalanffy model for R. cuneata in the Oder estuary
Line 318 Table 4. Comparison of von Bertalanffy equation parameters for different native and non-native populations of R. cuneata
The article lacks a broader environmental interpretation of the recorded size variation. What is the size of seawater intrusion and its spatial extent in the Oder estuary? What effect does this have on the obtained indices?
We agree with the above comment. In the discussion in section 4.1 Morphology, as well as in Figure 10 and Table 4, we presented how the parameters determining the growth rate of Rangia cuneata change in different environments.
Just before Table 4, a paragraph was added:
According to Wolnomiejski and Witek (2013) [60] Szczecin Lagoon connects with Pomeranian Bay, part of the Baltic Sea, through 3 straits: Peene, Świna and Dziwna. Based on data from 1998 – 2002, the average annual outflow of water from the Oder River to the Szczecin Lagoon is 18.38 km3. The largest outflow of water from the Szczecin Lagoon to the Pomeranian Bay takes place through the Świna Strait (78%), followed by Peene (14%) and Dziwna (8%). According to Wielgat (2002) [61], sea waters penetrate up to 100 km, and occasionally up to 160 km upstream from the mouth of the Oder. Thus, it can be considered that they penetrate into every part of the Szczecin Lagoon.
The authors point to the special role of water temperature, but there is no compilation of such data during the study period. The analysis of this parameter is too general in nature (Line 271-274).
We have included table 3 showing how the water temperature changed in several characteristic parts of Szczecin Lagoon in the years 2020-2023 and presented a short interpretation of this table:
We have included: In Table 3, the average temperatures of water in the Szczecin Lagoon in the vicinity of Stepnica, Nowe Warpno and Wolin from 12 November to 20 December in 2020-2022. Except for 12-18 November 2022, the analysis of 2022 data in table 4 did not show differences in average temperatures of water in various parts of the Szczecin Lagoon.
Table 3. Comparison of water temperatures on November 12 to December 20 in 2020-2022 [from: www.temperaturamorza.pl, access date: 21.08. 2023]
|
Date |
Stepnica |
Nowe Warpno |
Wolin |
||||||
|
2022 |
2021 |
2020 |
2022 |
2021 |
2020 |
2022 |
2021 |
2020 |
|
|
12 Nov |
11°C |
10°C |
10°C |
10°C |
10°C |
10°C |
11°C |
9°C |
10°C |
|
13 Nov |
10°C |
10°C |
10°C |
11°C |
10°C |
10°C |
11°C |
10°C |
10°C |
|
14 Nov |
11°C |
10°C |
10°C |
11°C |
10°C |
9°C |
11°C |
10°C |
9°C |
|
15 Nov |
11°C |
9°C |
9°C |
10°C |
8°C |
10°C |
11°C |
9°C |
10°C |
|
16 Nov |
11°C |
9°C |
10°C |
11°C |
8°C |
10°C |
11°C |
9°C |
9°C |
|
17 Nov |
11°C |
9°C |
10°C |
11°C |
9°C |
9°C |
11°C |
9°C |
9°C |
|
18 Nov |
11°C |
9°C |
9°C |
10°C |
9°C |
10°C |
10°C |
9°C |
10°C |
|
19 Nov |
10°C |
8°C |
10°C |
10°C |
8°C |
10°C |
10°C |
9°C |
10°C |
|
20 Nov |
10°C |
8°C |
9°C |
10°C |
8°C |
9°C |
10°C |
9°C |
9°C |
|
21 Nov |
10°C |
9°C |
9°C |
10°C |
9°C |
9°C |
10°C |
9°C |
9°C |
|
22 Nov |
9°C |
9°C |
9°C |
9°C |
9°C |
9°C |
9°C |
9°C |
9°C |
|
23 Nov |
9°C |
9°C |
9°C |
9°C |
8°C |
9°C |
9°C |
9°C |
9°C |
|
24 Nov |
9°C |
8°C |
9°C |
8°C |
8°C |
9°C |
9°C |
9°C |
9°C |
|
25 Nov |
8°C |
8°C |
8°C |
8°C |
8°C |
9°C |
9°C |
8°C |
9°C |
|
26 Nov |
8°C |
8°C |
8°C |
8°C |
8°C |
8°C |
8°C |
8°C |
8°C |
|
27 Nov |
8°C |
8°C |
8°C |
8°C |
7°C |
9°C |
8°C |
8°C |
8°C |
|
28 Nov |
8°C |
7°C |
8°C |
8°C |
8°C |
8°C |
8°C |
8°C |
8°C |
|
29 Nov |
8°C |
7°C |
8°C |
7°C |
8°C |
8°C |
8°C |
8°C |
8°C |
|
30 Nov |
8°C |
7°C |
9°C |
8°C |
7°C |
9°C |
8°C |
7°C |
9°C |
|
1 Dec |
7°C |
7°C |
8°C |
7°C |
7°C |
8°C |
7°C |
8°C |
7°C |
|
2 Dec |
7°C |
7°C |
7°C |
7°C |
7°C |
7°C |
7°C |
8°C |
7°C |
|
3 Dec |
6°C |
7°C |
7°C |
6°C |
7°C |
7°C |
6°C |
7°C |
7°C |
|
4 Dec |
6°C |
7°C |
7°C |
6°C |
7°C |
7°C |
7°C |
7°C |
7°C |
|
5 Dec |
6°C |
7°C |
7°C |
6°C |
7°C |
7°C |
6°C |
7°C |
7°C |
|
6 Dec |
6°C |
7°C |
7°C |
6°C |
7°C |
6°C |
6°C |
6°C |
7°C |
|
7 Dec |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
7°C |
6°C |
|
8 Dec |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
|
9 Dec |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
6°C |
|
10 Dec |
6°C |
5°C |
6°C |
6°C |
6°C |
6°C |
6°C |
5°C |
6°C |
|
11 Dec |
6°C |
5°C |
6°C |
5°C |
5°C |
6°C |
6°C |
6°C |
5°C |
|
12 Dec |
5°C |
5°C |
6°C |
5°C |
5°C |
5°C |
5°C |
5°C |
5°C |
|
13 Dec |
5°C |
5°C |
5°C |
5°C |
5°C |
5°C |
5°C |
5°C |
5°C |
|
14 Dec |
5°C |
5°C |
5°C |
5°C |
4°C |
5°C |
5°C |
5°C |
5°C |
|
15 Dec |
5°C |
4°C |
5°C |
5°C |
4°C |
5°C |
5°C |
5°C |
5°C |
|
16 Dec |
5°C |
5°C |
5°C |
4°C |
4°C |
5°C |
5°C |
4°C |
5°C |
|
17 Dec |
5°C |
5°C |
5°C |
4°C |
5°C |
5°C |
5°C |
5°C |
5°C |
|
18 Dec |
4°C |
4°C |
5°C |
4°C |
4°C |
5°C |
4°C |
5°C |
5°C |
|
19 Dec |
4°C |
4°C |
5°C |
4°C |
5°C |
5°C |
4°C |
5°C |
5°C |
|
20 Dec |
4°C |
4°C |
5°C |
4°C |
4°C |
5°C |
4°C |
5°C |
5°C |
Sentence: It 272 should be mentioned that water temperatures in the Szczecin Lagoon range from 0.5°C (in winter) to 26°C (in summer), with an annual average of approximately 11oC [56]. The article was cited as the source of the information: Verween, A.; Kerckhof, F.; Vincx, M.; Degraer, S. First European record of the invasive brackish water clam 512 Rangia cuneata (G.B. Sowerby I, 1831) (Mollusca: Bivalvia). Aquat Invasions 2006, 1, 198-203. 513 DOI:10.3391/ai.2006.1.4.1. The studies collated there concern the Harbour of Antwerp (Belgium), and hence cannot be referred to the Oder estuary. This should be changed.
Also, what is the role of the temperature of the Oder estuary (southern sites) on the results obtained compared to the northern sites?
We agree with the above comment. We have corrected the order of literature. In the literature is incorrectly set, position 56 is Verween et al. (2006). We intended it to be a chapter by Radziejewska and Schernewski (2008) about the Szczecin Lagoon. In the context of the relevant literature, temperatures in the Szczecin Lagoon are key data to explain the rate of growth and probably mortality of R. cuneata. In waters with higher temperatures, the growth of this species is faster (Wolfe and Petteway 1968, Fritz et al. 2022). In addition, in areas with low temperatures in winter (e.g. Vistula Lagoon and probably Szczecin Lagoon), there is a higher mortality rate (Kornijów et al. 2018).
The temperature of the mouth of the Oder has no influence on the results obtained in relation to the northern sites. Temperature is only relevant in a given area in relation to the local population.
Line 323-324. "This section is not mandatory but can be added to the manuscript if the discussion is unusually long or complex". This is an editorial annotation. It should be removed from the text.
We agree with the above comment. We have removed the above note from the text.
The conclusions section should address the question of further research. What research can or must be done?
We agree with the above comment, but we would like to leave our proposed conclusions that summarize the entire article and add:
The impact of this species on ecosystems in southern estuaries of the Baltic Sea still remains unknown. On the one hand, in new areas, R. cuneata may compete with native species for food and substrate, and on the other hand, it may be food for benthic fish.
Due to biological diversity between R. cuneata populations living in different conditions and its fairly recent spread in estuaries of the Baltic Sea, the study on the species should enable to determine the role of the clam in the new aquatic environment.
Reviewer 2 Report
See attached notes
Comments for author File: 
Comments.pdf
Mostly ok, consistency in location name required. Some instructions to authors still present.
Author Response
Reviewer 2
Thank you very much for your review. We have made every effort to introduce all the corrections that were detailed in the review. We believe that the changes proposed by the reviewer will increase the substantive value of the work.
Line 26. There is a serious formatting problem with the references. The numbers clearly do not refer to the correct references. This makes it very difficult to assess whether the references are relevant or whether the research has been correctly reported. Two cases on this page are of particular concern. The reference to Grabowski [3] can be located because there is a reference [17] listed where M. Grabowski is the first author. That paper is concerned with life history traits of gammarid Amphipoda but is cited as if its findings applied in general to all life forms (including the molluscs in this manuscript). That paper does not use the term “abortive reproduction” (which is nonsensical) but does refer to number of generations per year which is not the same as “reproduction (multiple times per year)”. Multiple breeding attempts (clutches) is not the same as multiple generations. That paper has been misquoted and is of questionable relevance. The second case is that of reference [6] in the context of two species (a barnacles and sandpiper) that have apparently found their niche in the Baltic and are thus not considered invasive. I cannot check the veracity of this citation because I could find no mention of these two species in the titles of any references listed. This may be a case of ‘second-hand citation’ but it is not possible for me to check if the original source really did establish (or maybe believe) that these two species no longer compete with native species.
We agree with the comments. By mistake, the literature was inserted in alphabetical order, and not according to the course of its appearance in the article. This has already been corrected.
We have changed citations from Grabowski et al. [3] on Colautti and MacIsaac (2004) [3].
Line 32-34 - the work of Janas and Kędzierska (2014) [6] was quoted here, who mention Mya arenaria and Amphibalanus improvisus as species that after more than 100 years of existence in the Baltic Sea have permanently entered the species composition of the Baltic macrozoobenthos.
We inserted: According to Colautti and MacIsaac [3], for the new territory to be colonised, a potentially invasive species must go through several stages: (i) it needs to be present within the transfer vector and actually be transferred to a new territory, (ii) individuals need to survive in the new environment, and (iii) reproduce.
Line 45. “this species was found … in the western part of the Baltic Sea in 2018”. I presume the Oder estuary is considered part of the western Baltic so had this species only been known from the area 4 years before your study? How, then, do you account for 6+ ages of the bivalve in your study?
Even if R. cuneata had been present earlier than when first reported, it is possible that the age distribution could be truncated, and not at equilibrium, simply due to the short time since colonisation. This could have important consequences for interpretation of your data so it is important to expand on the timing of arrival of this species in your study area, with more information on when it was first reported, how big the shells were when first reported, and whether surveys (with negative results) had been done before the first reports.
Yes, the expansion of R. cuneata in the Baltic Sea is westward and northward, starting from the Vistula Lagoon, where the species was recorded for the first time in the Baltic Sea. Panicz et al. (2022) indicate that the first record in the Pomeranian Bay took place in October 2018, from where, together with inflows from the Baltic Sea, the larvae could get to the Szczecin Lagoon. It should be remembered that the oldest individuals from the Pomeranian Bay were over 8 years old, so the age of 6 years obtained by us is quite plausible. In addition, although the first individuals in the Pomeranian Bay were caught in 2018, this does not exclude the possibility of R. cuneata occurring in earlier years.
Line 66. The introduction has failed to explain why changes in shape of a bivalve has any relevance to the threats posed by an invasive species, or to its management.
We agree with this remark. The shape of shells, especially in juveniles (less than 10 mm in length), is similar to the Limecola balthica mussel, native to the Baltic Sea, which may make it difficult to identify R. cuneata co-occurring with L. balthica.
We have inserted: Information about changes in width and height of a shell and the elongation and convexity indices can be used to distinguish alien and invasive Rangia cuneata from indigenous species, especially from Limecola balthica. In addition, in different populations of the species, changes in measurable features of enable to trace changes of shape in different environments as a consequence of adaptation to water in the new environment.
Line 73. Figure 1 is a map and does not depict the sampling device. The reference to the Figure is misplaced.
We agree with this remark. We moved the citation of Figure 1 in such a way that it clearly informs that it is a map showing sampling locations.
We added: Between 12 November and 20 December 2022, 504 specimens of Rangia cuneata were collected during sampling of benthic organisms in the Szczecin Lagoon (Oder estuary) (Figure 1) using a Van Veen grab sampler within a sample area of 0.1 m2
Line 75. How were the specimens “taxonomically identified”? Are there other similar bivalves that it could be confused with (especially juveniles)? What published resources and key characters were used to check identification? Since an aim of this paper appears to be to see if this bivalve changes its shape during invasion of a new location, was identification independent of shell shape?
Individuals were identified based on the characteristics of Rangia cuneata detailed in the work of Piechocki and Wawrzyniak-Wydrowska (2016) and then the identification of each individual was verified by Dr. Anna Łabędzki from the Jagiellonian University in Krakow. The most characteristic features of R. cuneata are a more massive and more convex shell, a prominent and forward-curved hump. According to Soloviova et al. (2019), R. cuneata can be confused with Limecola balthica, especially for specimens up to 10mm.
Line 82. Rangia cuneata (and Mactridae in general) are usually known (in English) as clams, not mussels (which generally refers to Mytilidae or other bivalves with byssal attachment).
We agree with this remark. We have replaced all instances of mussels with clams
Line 100. I think you mean Additional measurements, but it is not clear why these measurements are different from “standard measurements”. Were these additional measurements only carried out on some of the specimens? Were they analyse differently?
We agree with this remark. The measurements of clam shells were divided into two groups: basic, i.e. the length, width and height of the shell and its weight, as well as additional measurements. The latter is a novelty described in the article as no such measurements of R. cuneata shells have been done previously. Additional measurements were performed on all 504 specimens examined.
Line 108. Does wet weight include the shell? This is not clear.
We agree with this remark. The wet mass includes the shell with the viscera and soft tissues. Weighing was done after 15 minutes of drying the open individual on a paper towel.
We included: During measurements, individual specimens were weighed on a WPS 600/C/2 electronic scales (RadWag, Poland) to determine their wet weight (g) (including the shell with soft tissues with an accuracy of 0.01g To determine the condition of bivalves (condition index CI), soft tissues and shells were dried in an oven at 60°C for 72 h to obtain dry weight [24].
Line 111. “I checked with other works”. You must cite the references to these other works.
We have decided to remove this sentence from the manuscript.
Line numbers are missing from section 2.3, but in the paragraph above equation (4) you state “true annuli were agreed upon”. Who agreed? How many independent assessors were there? Was it ‘blind’ agreement or open consensus?
We agree with this remark. Unfortunately, the file that is available on the MDPI website does not show the numbering of the lines at all and we could not manage to include this numbering in the manuscript.
As mentioned in the methodology, age was assessed independently by two authors who did not know the outcome determined by the other investigator. In case of discrepancies, consensus was reached to agree on the age of each individual.
Equation 4. This is the von Bertalanffy growth equation. You refer (by name) to this function elsewhere in the manuscript but should name it here for the convenience of the reader.
We agree with this remark. Added von Bertalanffy before the word formula (before formula #4).
There is a minor problem with equation (5). The inverse of equation (4) requires the addition of t0 as the final step. Your estimate of t0 is -0.47 so, by this method, your estimate of Amax will be nearly half a year too high.
For equation 5, we followed the manuscripts of Ziuganov et al. (1994) and Czerniejewski et al. (2021).
But there are more serious problems with this part of the methods. The von Bertalanffy growth curve (at least when fitted with Normal errors) models mean length at age. L∞ is the estimated mean (not maximum) length at infinite age. It appears that you have solved equation (5) using the maximum length recorded in your sample (43.2 mm) which happens to be from an individual aged 5 years. By what logic do you expect the maximum length at age 5 to be a good estimate of the mean length at the maximum age? Note that it is common in applications of the von Bertalanffy model (though not in your case) to find individuals that are larger than L∞. This is perfectly reasonable since L∞ is a mean length, but in such cases your equation (5) would fail.
In the von Bertalanffy equation, according to Ogel (2016) [16] on page 213, there is a sentence: “First, L∞ is the maximum mean length; it is not the maximum length of an individual. In otherwords, it is possible that an individual fish is longer than L∞” therefore it can be considered that this is the maximum average length of individuals in the sample, so it may happen that individual fish will be longer than L∞.
A further problem is that your von Bertalanffy curve in Fig. 2a is clearly under-fitting the data for the oldest two age classes. Extrapolating the curve to estimate the maximum age is clearly not appropriate with such a poor fit. Perhaps a different growth model might fit your data better. And finally, inverting the growth model to predict age from length is not statistically valid because the model is fitted by minimising error in the lengths. If you want a model to predict age from a particular length, then you should fit the model A ~ L (i.e. equation 5) and use that to make your extrapolation (with appropriate confidence intervals on Amax). Given the many problems with estimating Amax from Lmax, I suggest you consider the method commonly used in fisheries modelling which is extrapolation of the ‘catch curve’ to the survival of some arbitrary proportion (say 1%) of the cohort. However, catch curve analysis relies on the population being at equilibrium which may not be the case (see my earlier comment). I also wonder why you wish to estimate the maximum age. Is it an attempt to estimate the time of first arrival of the species in your study area? If that is the case then looking for truncation of the survivorship (catch) curve would be more useful than an extrapolated upper age limit (with equation 5) that may not have had time to be achieved
After consulting with Prof. Derek H. Ogle, a specialist in the analysis of the von Bertalanffy curve, we learned that with a large variability of the number of individual age classes, as is the case in our research, length classes represented by a large number of individuals will be better modelled by the trend line drawn by the model von Bertalanffy. For this reason, we decided not to carry out the "Catch curve analysis" suggested by the Reviewer.
We decided to analyze the maximum age after a thorough study of articles by, among others, Kornijów et al. (2018) stating that Rangia mortality occurs at the end of the winter period, when the condition of mussels decreases, and that the oldest and youngest individuals are exposed to them.
The analytical/statistical methods are not well explained. In the section between equations (4) and (5) you list several R packages but don’t state what they were used for or exactly what models were fitted. In the case of the von Bertalanffy growth curve, did you fit the model by least squares (assuming Normal errors) or maximum likelihood (to allow for other error distributions such as lognormal)?
Most of the R packages listed are utilities for data preparation or display and one (nnnet) is for constructing neural networks – but I don’t see that technique used in your manuscript. Could it be that this list of R packages was pasted from some other work without checking if all were relevant? You should state clearly what each package was used for and delete those not relevant.
We agree with this remark. We added short information about what the individual R packages were used for. When calculating the growth parameters using the von Bertalanffy equation, we used the least squares method.
Alternative names for the location Oder/Odra are used. It is acceptable to explain (in the Introduction) that there are alternative names in different languages but throughout the manuscript you must consistently use one version
We agree with this remark. We standardized all the names and decided on the word Oder, we also put the word Oder instead of Odra in figure 10 .
Section 2.4. Some aspects of statistical analysis have already been covered. All could be usefully collected here in one place. As well as R packages you now mention Statistica but once again, do not explain what analyses you performed with that software nor why it could not be achieved with the various R packages. Levene’s test is for assessing the equality of variances for two or more groups (as in ANOVA) but all your analyses seem to be regression models. Homoscedasticity is still important in regression models but requires different methods to assess. Levene’s test is not appropriate.
We agree with this remark. We mainly used different types of regression. We added the Breusch-Pagan test for heteroscedasticity.
In Statistica, graph 10 was made, which shows the relationship between φ′ and P and the average water temperature in various places where the samples were taken. The Statistica 13.0 program was used here due to the greater possibility of calculating and editing the charts after their creation (e.g. overlapping the names of research stations.
Table 2. The slope parameters for the last two models are very high compared to all other allometric relationships. The two variables involved AAAM and PAMP both measure the distance between some shell feature and the margin. Is this extreme allometry because the feature moves with the expanding shell or is it being left behind as the shell grows. Some interpretation would be useful here.
PAMP and AAAM values – based on the analysis of the data (correlation between shell length L and PAMP and AAAM), it can be concluded that this feature moves with the length of the shell.
We have inserted: The probable cause of the high values of the slope parameter for Log AAAM ~ Log L and Log PAMP ~ Log L is the rapid increase in the AAAM and PAMP parameters relative to the shell length. The AAAM and PAMP distances characterise, in a sense, the bulging of the shell.
Line 176. Changes in slopes within age classes can be “observed” in Fig. 7 but can it be justified statistically? A test for uniformity of slopes within an ANCOVA model would be useful here. The change of slopes in figure 8 is less convincing. Is there a significant overall slope in these graphs? It would be rather surprising if there was an increase in elongation within age classes but not between.
We agree with this remark. In the new version of the manuscript, we performed the ANCOVA test for the data in Figure 7. There seems to be a pattern here - before puberty (which occurs at the age of 2 years), there is a faster increase in width than in length (probably due to preparation for reproduction) , in the period of reaching maturity, the Elongation Index does not change with the length of the shell (it can be assumed that there is a stable and even increase in the length and height of the shell), and after reaching maturity, the growth of the shell width in relation to its length slows down.
We added: The ANCOVA analysis shows statistically significant differences between age classes relative to L: Elongation index (for Elongation index p-value: 0.000000002627, and for age: p-value: 2.2e-16, both values are statistically significant at the alpha level of 0.001).
Do these bivalves reset their shape at the start of each age class? If so, that would be interesting. In what season do your age classes begin? Is the increase in elongation due to differences in summer growth and winter growth?
There seems to be no "reset" of the shape in each age group, which can be seen in Figures 7 and 8. The shape of the shell, determined by the Elongation and Convexity index, changes continuously by adding another annual ring on the edge.
The available literature shows that in most species of mussels, including Rangia cuneata, the process of building another annual ring takes place in mussels in late spring before the reproductive period and is probably a response to the growth slowdown in winter.
In the winter, mussels in the continental climate zone do not grow due to the fact that at low water temperatures the efficiency of filtration decreases very rapidly. This causes deterioration of the condition of mussels and may be the cause of their mortality after the winter period.
Line 189. No, those significance tests are not at all interesting. The null hypotheses of L∞ = 0 and and K = 0 imply no size and no growth. Of course these bivalves have non-zero length and they grow. The non-zero t0 is also likely due to the within-year calibration of your aging method rather than any biologically interesting process. You have not explained at what stage in each year since ‘birth’ the growth rings are formed.
The coefficient K indicates the rate at which the curve built by the von Bertalanffy model tends to the value of L∞. In no case have we put H0 saying that K=0 and L∞=0. The significance level of each parameter of the von Bertalanffy equation is important in assessing the suitability for use in the model.
Line 195. The growth pattern you describe is a necessary consequence of the von Bertalanffy growth model but you have not properly assessed whether this model is a good fit to your data.
We agree with this remark. We have added information on model fit
We have inserted: The model presented in Figures 2a, b and c well illustrates the growth rate in length classes where many individuals were observed (age groups 1-4), and less well illustrates the growth rate of individuals from less numerous age classes, which, as in the case of age group 6, are greater than predicted by the model.
When I look at figure 2 (and ignore the fitted lines) I see slow growth between age classes 3 and 4 followed by a return to faster, almost linear, growth. Perhaps you should check the climatic conditions of those particular years to see if there was some event to temporarily slow the growth. But since your sample is from a very narrow time range (38 days), you are effectively dealing with one cohort in each age class, so the growth check could have been in any year 3-6 years before your sampling date.
We agree with this remark. Perhaps taking many samples during the year would show statistically significant differences between these research seasons. In the newly added table 3, we showed how the water temperature changes during the sampling period (12 November – 20 December 2022 and we have attached data from this period of 2020-2022) and a systematic decrease in water temperature.
Section 4.1. The first part of your discussion is pointless if you do not compare age structures of the populations you are comparing. The Oder estuary shells maybe smaller because they are a younger population. This could be because of the particular habitat (e.g. depth, sediment type) you sampled or because this is a recently established population that has not had time to mature large individuals. Another complication is the possibility of size selective harvesting in some populations. It is age-specific sizes that you should be comparing.
We agree with this remark. In section 4.1 we discuss our results with data from around the world and this seems to be a necessary part, so we do not want to remove it. In the future, we plan to conduct research that will answer the reviewer's questions.
Line 309. The range of parameter values in different populations of this species is rather large. It would be useful to provide confidence intervals in this table, or at least give sample sizes to indicate which studies might be more reliable. A graph comparing the fitted curves from the 5 populations could also be considered. The three parameters of the von Bertalanffy growth model are strongly correlated so some combinations of very different parameters (especially for K and L∞) can generate rather similar curves over a limited age range. This is the case for 3 of the populations (for ages 1-6) but the other two are very different curves (see figure below). This would be worth exploring and discussing further. The most useful thing you could do with these data is to establish whether there are statistically significant differences in the growth models of the different populations. Because of the strong correlations between the parameters of the model it is no use testing for differences of individual parameters (such as by comparing confidence intervals), you need to test for the effect of a population factor on the overall von Bertalanffy model.
We agree with this remark. We added information about the number of samples that we used to build Fig. 10.
