Impacts of Recreational Angling on Fish Population Recovery after a Commercial Fishing Ban

Round 1

Reviewer 1 Report

This paper has well studied the impacts of recreational angling on fish population recovery after a commercial fishing ban, which is important in managing the recreational fishery in Kaunas Reservoir, Lithuania. I would recommend publication after some revisions.

 

1, Because the title is about the impacts of a commercial fishing ban, I would expect the CPUEs before and after the ban explicitly in the RESULTS, e.g. as a new table.

2, In ABSTRACT, I would expect some explicit technical details (i.e. numbers).

3, In KEY WORDS, “modelling” can be replaced by “Kaunas Reservoir, Lithuania”.

4, In Figure 2, I would expect the unit for the “scientific CPUE”. For the recreational catches, I would expect the actual data, instead of “indicative trendline”, so as to provide more information.

5, In Table 1. “Estimated recreational catches” should be “Estimated recreational CPUE”. The year of this table should be clarified.

Author Response

Reviewer 1, general comments:

This paper has well studied the impacts of recreational angling on fish population recovery after a commercial fishing ban, which is important in managing the recreational fishery in Kaunas Reservoir, Lithuania. I would recommend publication after some revisions.

 

Authors’ response:

We thank the reviewer for their comments and suggestions and have endeavoured to accommodate them in our revision. Our response to the reviewer’s specific comments is itemised in the table below.

 

Specific comments:

Reviewer 1 comments	Authors’ response
1, Because the title is about the impacts of a commercial fishing ban, I would expect the CPUEs before and after the ban explicitly in the RESULTS, e.g. as a new table.	New supplementary table (A.5.) with scientific monitoring standardised CPUE time series was added.
2, In ABSTRACT, I would expect some explicit technical details (i.e. numbers).	Thank you, we added some details, while aiming to keep to the word limit.
3, In KEY WORDS, “modelling” can be replaced by “Kaunas Reservoir, Lithuania”.	Done
4, In Figure 2, I would expect the unit for the “scientific CPUE”. For the recreational catches, I would expect the actual data, instead of “indicative trendline”, so as to provide more information.	The figure shows relative change in all values, which are scaled to the maximum value, therefore the y axis is on relative unitless scale. Also, since CPUE is standardised we do not have units, because values are extracted as yearly deviations. However, to make clear that CPUE is in biomass and not numbers we indicate it in the figure caption, methods and results.
5, In Table 1. “Estimated recreational catches” should be “Estimated recreational CPUE”. The year of this table should be clarified.	Fixed, thank you.

 

Reviewer 2 Report

In this ms, the authors presented the cpue standardization results on several fish species in a reservoir based on both a fishery-independent survey and a creel survey. Based on the results, the authors show that after the commercial fishing ban, the recovery rate of some species depends upon fishing pressure from the recreational sector. The text is relatively well written. But the overall quality and the manuscript's organization can be further improved. See below for further details.

 

lines 133: Why there is no spatial factor in the formula? From the map, I assume the sampling locations are not randomly chosen but fixed in place across years, right? In this case, it makes sense to have a spatial factor. Do all the species distribute evenly in this reservoir?

 

lines 137-140: How did you eliminate the variables? Did you use forward/backward selection or something else?

 

lines 162: Somewhere around here, provide some more information on the fishing rules. Such as, are there daily limits, species-specific rules, catch and release rules, or size limits? Are these rules regularly enforced? These aspects are essential for this study.

 

Figure 2: In the capture, write out the unit of catch. It is very easy to mistake t for number 1 in the figure.

 

Also, why only the upper ranges are plotted? what about the lower ranges?

 

Why use 80% confidence interval? Please keep the confidence intervals consistent at 95% as in all other places.

 

Why there is missing data for the latter 3 figures between 1995 and 2000?

 

What do the cpue trend lines from the recreational sector look like? Do they follow the independent survey cpue?

 

The assumption that the recreational fishing effort steadily decrease to 50% of this current level in 1990 is not necessary. It may not be true. Probably it is not true. It is enough to show that for some fishes, the catch is high in the recreational sector.

 

Table 1: Which test does the p refer to?

 

Table 2: I couldn't find table A.5. Are you referring to table A.4 instead?

 

Check the labeling of all the tables.

 

In this table, the min and max are for different seasons. Judging from the caption alone, one would think these are for different years. Also, is it true that those seasons with an infinite catch have no recorded catch for those species? In that case, those season x species combinations should be omitted.

Author Response

Reviewer 2, general comments:

In this ms, the authors presented the cpue standardization results on several fish species in a reservoir based on both a fishery-independent survey and a creel survey. Based on the results, the authors show that after the commercial fishing ban, the recovery rate of some species depends upon fishing pressure from the recreational sector. The text is relatively well written. But the overall quality and the manuscript's organization can be further improved. See below for further details.

 

Authors’ response:

We agree with the reviewer’s comments and have endeavoured to accommodate them in our revision.

Our response to the reviewer’s specific comments is itemised in the table below.

 

Specific comments:

Reviewer 2 comments	Authors’ response
lines 133: Why there is no spatial factor in the formula? From the map, I assume the sampling locations are not randomly chosen but fixed in place across years, right? In this case, it makes sense to have a spatial factor. Do all the species distribute evenly in this reservoir?	Location was actually included in the CPUE standartisation; it was just missing in the equation. Thank you for noticing this The error is now fixed (ca lines 140).
lines 137-140: How did you eliminate the variables? Did you use forward/backward selection or something else?	We used backward selection. This was explained in lines 148-150  (“started using the full model with all explanatory variables, and then the parameter number was reduced based on Akaike’s Information Criterion”), but we now clarify it further (line 151, track changes mode).
lines 162: Somewhere around here, provide some more information on the fishing rules. Such as, are there daily limits, species-specific rules, catch and release rules, or size limits? Are these rules regularly enforced? These aspects are essential for this study.	Done.
Figure 2: In the capture, write out the unit of catch. It is very easy to mistake t for number 1 in the figure.	The figure is now modified. Also we corrected a small mistake that we found in the back-transformation of standardised scientific CPUE trend (from log scale to linear scale). This has modified the figure slightly and filled the missing areas which were previously back-transformed into to infinity values and therefore not shown on the plot. The trends and analyses result remain the same.
Also, why only the upper ranges are plotted? what about the lower ranges?	The uncertainty ranges for scientific CPUE are symmetrical (up and down) but only upper ranges are plotted, because lower ranges should be bound by zero (CPUE cannot be negative). The statistical analysis is done on a log scale with a tweedie distribution, so in the analysis itself this is not a problem. It is only when the means and error bars are back-transformed to the linear scale that the issue of plotting arises. This explanation about the lower bounds of CPUE trend is now added to the caption of the figure.
Why use 80% confidence interval? Please keep the confidence intervals consistent at 95% as in all other places.	The 80% intervals reflect Bayesian posterior probability limits and not ‘traditional’ confidence intervals from frequentist statistical model. Because Bayesian posterior probabilities are usually much wider than frequentist confidence intervals, in the published Fisheries Research paper we used 80% probability ranges of angling effort in Kaunas WR. Therefore, we use the same ranges also in this study. We now added an explanation about it in the methods (line 399-402, caption of Table 2) and use term uncertainty range in the table captions where both uncertainty values (from frequentist and Bayesian models) are combined Note, this combined uncertainty is only relevant for the estimation of the annual recreational catch (not recreational CPUE), such as in Table 2 and Figure 2.
Why there is missing data for the latter 3 figures between 1995 and 2000?	The missing data was related to the back-transformation of yearly deviation from log space into the linear space (as explained above), and this is now fixed. There are no more missing years in the figures.
What do the cpue trend lines from the recreational sector look like? Do they follow the independent survey cpue?	From our individual surveys conducted in 2016, 2017, 2020 and 2021, we did not see trends in recreational catch CPUE. This means there were no significant annual trends in this data set, but this most likely reflects the large inherent variation in recreational CPUE and the fact that we only had limited number of surveys per year. Whenever year was identified as an important variable for individual species recreational CPUE, annual deviations can be seen in the full model parameters shown in Figure A5. However, in many cases they are fluctuating randomly and there are no clear trends; many are not significantly different from 0 (except for one out of four studied years). Unfortunately, we do not have long term information on recreational catches and we cannot estimate longer recreational CPUE trends. We now added more explanation about this in lines 356-359 (track changes mode).
The assumption that the recreational fishing effort steadily decrease to 50% of this current level in 1990 is not necessary. It may not be true. Probably it is not true. It is enough to show that for some fishes, the catch is high in the recreational sector.	We agree that it is not necessary, however we feel that it is still important to provide at least approximate estimates of how recreational catch could have changed. This assumption is made on the only data available to assess recreational effort – sales of annual recreational licences. While the data is limited, it is still valid data, used in many studies of recreational effort, and suggests that recreational effort has increased. This is also supported by anecdotal evidence and general observations. Therefore we prefer to keep the line, but note that we do not add any confidence ranges around it, to show that the lines is just a broad assumption.
Table 1: Which test does the p refer to?	The p value refers to the variables that were significantly different from zero as identified through t-statistics in general linear model analysis. Thank you for detecting this missing explanation, it is now added to the Table 1.
Table 2: I couldn't find table A.5. Are you referring to table A.4 instead?	Apologies for this confusion with tables, this is now fixed
Check the labeling of all the tables.	Thank you, this is fixed
In this table, the min and max are for different seasons. Judging from the caption alone, one would think these are for different years. Also, is it true that those seasons with an infinite catch have no recorded catch for those species? In that case, those season x species combinations should be omitted.	The table reports annual average catch over the 2016-2020 and the min-max values refer to the annual total catch, not the season. Although the model included seasons for many species, the predicted average catch accounts for this difference, while making the final estimate. You are correct that seasons with infinite catch estimates were omitted from the annual mean predictions (clarified in lines 375-376, track changes mode).

 

Reviewer 3 Report

This is a nice paper addressing the impact of recreational angling in freshwater fish populations, a topic that it is seldom explored also given the inherent difficulties in recording recreational fishing efforts and catches. Nevertheless, there are evidence on how recreational fisheries can negatively impact freshwater fishes and further studies are needed to improve recreational fishing management and make it more sustainable. Therefore, this study contributes evidencing how, for some species, recreational angling harvest is higher than the commercial one and aligns with a decrease in the species catch per unit effort, also discussing some management suggestions. The paper is generally well written, the introduction gives a complete picture of the problematic, the methods used are generally fine and take advantage of a commercial fishing ban in the study area that allowed appreciating the impacts of recreational fishing only. The main results are clear and well discussed. I only have two main concerns. First, some material/results are missing (i.e, Table A. 5 is not provided) or reference to supplementary material is confused (see specific comments for further details). The second concern is about the two methodologies used to infer the recreational fishing efforts: for 2020, the authors combined as per their previous publication, drone-based assessment and fishfinder device record. For other years they inferred it from fishing licence released by the Ministry. I am concerned about how the authors put together these data, whether they are comparable or not (see the specific comment below). I wonder if it is not better, for consistency, to use only the licence method. This method of course has limitations but at least gives comparable output across years. Finally, fish population dynamics may have been influenced by evolution over time in habitat characteristics (e.g., limnological parameters variation), which has not been considered. Thus, ideally an analysis of the evolution of the limnological parameters of the reservoir would be necessary but if not possible, this should be added as a limitation in the appropriate section already present in the discussion. Apart from these concerns, I think the paper is suitable for publication in Fishes, pending some further correction listed below as specific comments.

Line 30 change “population” with “species”

Line 65 and the rest of the manuscript: when reporting the scientific name of the species, remove brackets (they are not necessary) and report the authority of the species at the first mention.

Line 73 Check the font size

Line 77 Correct the verb to “has supported”

Line 89-90 I would rephrase as “could be explained by differential recreational harvesting of fishes” since you did not assess the mortality rate.

Fig.1 The resolution of this picture is a bit poor.

Line 140 There are two Table A.1 in the supplementary material. The first of which offers a list of variables categories, and it is not even a table.

Line 144 correct to “back-transformed”

Line 153 Table A.2 is called here to show the results of the sensitivity analysis. However, this table shows the questionnaire used for the angler survey. So, make sure to quote the right supplementary material and correct the number of each table.

Line 197-207 But are the two outputs of the different methods used (recreational angling licence and drone), actually comparable? It is not very clear how the data obtained with these two different approaches have been combined. If relevant, a comparison between efforts obtained using licence only and drone-fish finder device for the years where both are available may be useful.

Table A.3 Please, use as decimal separator “.” and not comma.

Line 242 Table A. 4 is citated, but that table does not contain the parameters of the full and seasonal models, respectively. So again, make sure to quote the right material where these results are presented. I think you should quote Fig. A.5 but be aware that in the caption, a Table sX is quoted which appear to be missing.

Line 508 correct to “they are also likely”

Line 515 Remove comma after “management”

Author Response

Reviewer 3, general comments:

This is a nice paper addressing the impact of recreational angling in freshwater fish populations, a topic that it is seldom explored also given the inherent difficulties in recording recreational fishing efforts and catches. Nevertheless, there are evidence on how recreational fisheries can negatively impact freshwater fishes and further studies are needed to improve recreational fishing management and make it more sustainable. Therefore, this study contributes evidencing how, for some species, recreational angling harvest is higher than the commercial one and aligns with a decrease in the species catch per unit effort, also discussing some management suggestions. The paper is generally well written, the introduction gives a complete picture of the problematic, the methods used are generally fine and take advantage of a commercial fishing ban in the study area that allowed appreciating the impacts of recreational fishing only. The main results are clear and well discussed.

Authors’ response:

Thank you for a thorough review. We agree with reviewer’s general comments and greatly appreciate the suggested corrections and guidance.

 

Reviewer 3:

I only have two main concerns. First, some material/results are missing (i.e, Table A. 5 is not provided) or reference to supplementary material is confused (see specific comments for further details).

Authors’ response:

Thank you and apologies for the confusion. We believe this is all fixed now and we checked all the supplementary materials carefully.

 

Reviewer 3:

The second concern is about the two methodologies used to infer the recreational fishing efforts: for 2020, the authors combined as per their previous publication, drone-based assessment and fishfinder device record. For other years they inferred it from fishing licence released by the Ministry. I am concerned about how the authors put together these data, whether they are comparable or not (see the specific comment below). I wonder if it is not better, for consistency, to use only the licence method. This method of course has limitations but at least gives comparable output across years.

Authors’ response:

This study of recreational catches and their comparison with commercial fishing catches were possible because we conducted detailed assessment of recreational effort. Recreational licences are sold on a country level, not on a water reservoir level. Therefore it is impossible to know, based on licence sales alone, how much effort occurred in Kaunas WR. Limiting our analysis to licence method only would negate majority of our results and findings. Therefore the combination of two methods to assess recreational effort was essential. Indeed, to combine the two methods, we had to make an assumption that the spatial distribution of anglers remained approximately similar over the last five years and that relative changes in licence sales reflect relative effort changes in Kaunas WR. This is a limitation, but we still have a lot more data on effort than based on licence sales alone. For example, earlier guesses on recreational annual recreational effort in Kaunas WR (based on country wide licence sales) were in the range of 10 thousand trips per year. In our new assessments (Dainys et al. 2022) we show that the number of trips is 10 times higher (ca 110 thousand trips per year).

 

Reviewer 3:

Finally, fish population dynamics may have been influenced by evolution over time in habitat characteristics (e.g., limnological parameters variation), which has not been considered. Thus, ideally an analysis of the evolution of the limnological parameters of the reservoir would be necessary but if not possible, this should be added as a limitation in the appropriate section already present in the discussion. Apart from these concerns, I think the paper is suitable for publication in Fishes, pending some further correction listed below as specific comments.

Authors’ response:

Unfortunately we cannot address questions about possible evolutionary changes and influences in the current study as this would require an extensive data collection and analysis and a separate publication. We do agree this should be considered, and added this caution to the discussion (lines 567-574, track changes mode).

Our response to each of the specific comments is provided in the table below.

 

Specific comments

Reviewer 3 comments	Authors’ response
Line 30 change “population” with “species”	Corrected
Line 65 and the rest of the manuscript: when reporting the scientific name of the species, remove brackets (they are not necessary) and report the authority of the species at the first mention.	Brackets were removed and authority of the species added throughout the text.
Line 73 Check the font size	Corrected
Line 77 Correct the verb to “has supported”	“has” was added before “supported”
Line 89-90 I would rephrase as “could be explained by differential recreational harvesting of fishes” since you did not assess the mortality rate.	Done
Fig.1 The resolution of this picture is a bit poor.	Agree, resolution of the image was increased.
Line 140 There are two Table A.1 in the supplementary material. The first of which offers a list of variables categories, and it is not even a table.	Apologies for this confusion with tables, this is now fixed
Line 144 correct to “back-transformed”	Corrected
Line 153 Table A.2 is called here to show the results of the sensitivity analysis. However, this table shows the questionnaire used for the angler survey. So, make sure to quote the right supplementary material and correct the number of each table.	Apologies, we believe all confusion  is now fixed
Line 197-207 But are the two outputs of the different methods used (recreational angling licence and drone), actually comparable? It is not very clear how the data obtained with these two different approaches have been combined. If relevant, a comparison between efforts obtained using licence only and drone-fish finder device for the years where both are available may be useful.	The drone and fishfinder data, published in detail in Dainys et al. 2022 is only available for 2020-2021. As we mentioned earlier the licence data is collected at country level and cannot be, by itself, used to assess effort in a specific water reservoir or lake. Therefore we need to use a combination of both, as they reflect different things. This is now explained further in lines 236-243 (track changes mode).
Table A.3 Please, use as decimal separator “.” and not comma.	Corrected
Line 242 Table A. 4 is citated, but that table does not contain the parameters of the full and seasonal models, respectively. So again, make sure to quote the right material where these results are presented. I think you should quote Fig. A.5 but be aware that in the caption, a Table sX is quoted which appear to be missing.	We believe this is now fixed, apologies again.
Line 508 correct to “they are also likely”	Corrected
Line 515 Remove comma after “management”	Comma was removed

 

Round 2

Reviewer 2 Report

The authors have done a good job in revising the ms. I have no further comments except that there exist some occasional typos in the peer-review-v2.pdf file, for example, line 569, but I think the proofreader will correct them before publication. Again, well done.

Author Response

Thank you, fixed.