Roe Deer, Lithuania’s Smallest and Most Abundant Cervid

Round 1

Reviewer 1 Report (Previous Reviewer 3)

I have no further comments; the author's answers are satisfactory to me.

Author Response

Thank you

Reviewer 2 Report (Previous Reviewer 1)

In my opinion manuscript has been significantly improved.

Author Response

Thank you

Reviewer 3 Report (New Reviewer)

I have a number of minor comments that require addressing:

1. Avoid using significant when you mean substantial

2. Line 88. Have you confirmed the theses did not contain original information, or was it an assumption.

3. Line 201. Clarify what you mean here. "their numbers" technically refers to wolves and I think you mean roe deer. But if roe deer numbers remained high, then how can poaching have a "significant influence"?

4. line 216 delete "decrease of" because you can't negatively impact a decrease.

5. line 238.I would suggest you say "were reported hunted"

6. line 247/248. This drop in hunting seems relatively small given the covid restrictions. Can you say how this impacted hunting effort?

7. Line 345. Define what you men by "critical for the species" as this would imply it is necessary for the species survival and I think you mean the opposite.

8. line 351 perhaps replace agrocenoses with a description as it is not a well used word.

9. Line 386. It would be good to give details of where the Pannonian region is.

10. line 493 replace best with big.

11. line 543 correct spelling of densities.

12. line 596 correct "leacve spacefor" to leave space for"

Overall quality of English is excellent, with just some minor changes required to make this easier to read.

Author Response

Rev#3 comments and answers

I have a number of minor comments that require addressing:

Comment 1. Avoid using significant when you mean substantial

Answer: thank you, changes done

Comment 2. Line 88. Have you confirmed the theses did not contain original information, or was it an assumption.

Answer: no, not an assumption. Theses were downloaded and analysed. I added explanation to the text.

Comment 3. Line 201. Clarify what you mean here. "their numbers" technically refers to wolves and I think you mean roe deer. But if roe deer numbers remained high, then how can poaching have a "significant influence"?

Answer: Apologies, maybe it was not clear, that wolf numbers and poaching had an influence to the period mentioned in the first sentence. I changed it to: The next decade, 1948–1958, was characterized by a slow increase in roe deer numbers (CAGR = 0.084): despite the extermination, numbers of wolves remained high, and poaching still had a significant influence.

Comment 4. line 216 delete "decrease of" because you can't negatively impact a decrease.

Answer: corrected.

Comment 5. line 238.I would suggest you say "were reported hunted"

Answer: corrected as adviced.

Comment 6. line 247/248. This drop in hunting seems relatively small given the covid restrictions. Can you say how this impacted hunting effort?

Answer: No data on changes of hunting effort during COVID are available. Therefore, we changed sentence as: There was a drop in roe deer hunting bag size and proportion of the population hunted from 2020 to 2022 (see Figures 3 and 4), however, there are no data on changes in hunting effort.

Comment 7. Line 345. Define what you men by "critical for the species" as this would imply it is necessary for the species survival and I think you mean the opposite.

Answer: apologies, I used word construction from the cited source. Now corrected. Hope you will acceot wording: Snow cover exceeding 40 cm was considered detrimental to the species survival.

Comment 8. line 351 perhaps replace agrocenoses with a description as it is not a well used word.

Answer: changed to “agricultural habitats”

Comment 9. Line 386. It would be good to give details of where the Pannonian region is.

Answer: authors cited mean “Pannonian mixed forests ecoregion in Central south-eastern Europe (Hungary)”, so Hungary was inserted to explain sample origin.

Comment 10. line 493 replace best with big.

Answer: changed to “main”

Comment 11. line 543 correct spelling of densities.

Answer: corrected

Comment 12. line 596 correct "leacve spacefor" to leave space for"

Answer: corrected

Thank you for all your comments.

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

The reviewed manuscript (forests-2909600) "Roe deer, Lithuania's smallest and most abundant cervid" raises a very current issue. A review of 114 scientific and other publications on the smallest and most abundant deer in Lithuania, the roe deer (Capreolus capreolus) in Lithuania is presented, along with an analysis based on compound annual population growth rates of population numbers, hunting bag and roadkills. This review also covers the species' history in the country from the last glaciation onwards, the changes in numbers from 1934 to 2023, hunting bag changes from 1965 to 2022, roadkill numbers from 2002 to 2022, data on roe deer reproduction, habitat selection, genetic diversity, pathogens and damage to forest stands. It also provides an overview of species management and selection for trophies. Therefore, the analysed topic research is very important.

The research material used was selected appropriately, as well as the statistical methods used to analyze it. Own results and their discussion in the Discussion are properly described and were compared with previous studies by other authors. I have no objections to the substantive and editorial value reviewed article.

Author Response

Reviewer 1 comments and answers

Comment: The reviewed manuscript (forests-2909600) "Roe deer, Lithuania's smallest and most abundant cervid" raises a very current issue. A review of 114 scientific and other publications on the smallest and most abundant deer in Lithuania, the roe deer (Capreolus capreolus) in Lithuania is presented, along with an analysis based on compound annual population growth rates of population numbers, hunting bag and roadkills. This review also covers the species' history in the country from the last glaciation onwards, the changes in numbers from 1934 to 2023, hunting bag changes from 1965 to 2022, roadkill numbers from 2002 to 2022, data on roe deer reproduction, habitat selection, genetic diversity, pathogens and damage to forest stands. It also provides an overview of species management and selection for trophies. Therefore, the analysed topic research is very important.

The research material used was selected appropriately, as well as the statistical methods used to analyze it. Own results and their discussion in the Discussion are properly described and were compared with previous studies by other authors. I have no objections to the substantive and editorial value reviewed article.

Answer: Thank you for your feedback and positive review of my manuscript. I appreciate acknowledgment of the importance of our research topic.

Reviewer 2 Report

Figure 1. I understand that this map presents current borders of Lithuania, are data presented in figure 2 from 1934-1940  relevant to this area?

I don’t understand how roe deer were counted. I found only information about snow-tracking and roe deer observations. Generally counting of free living animals is extremely difficult. Different methods give different results. It’s impossible to count animals precisely. Even using the most advanced, modern methods, the error of evaluation exceeds 25%. How is it  possible that population grew so quickly between 2016-2020? I am afraid that presented data may not have any scientific value, being probably only imagination of hunters. It’s not clear if it is data from the whole country or only from hunting grounds? The same problem is with data about hunting bag. It is official data from hunting reports but real harvest is unknown. Unregistered hunting and poaching is a common practice in the whole Europe. Hunting bag might be a density trend indicator, but only under some circumstances: the numbers of hunters and their hunting activity being stable during the whole analysed period, length of hunting season and methods being the same, shooting is unlimited by management plan.

Also data about roadkill are questionable. Roe deer is relatively small and probably only small fraction of killed animals are officially reported.

In Lines 345-349 there is information about roe deer pellet group counting, the biggest problem with this method is fact, that detection rate and accumulation period of pellet groups is different in different habitats. Because of this, its impossible to study roe deer habitat selection using this method.

In section 3.9, how the forester distinguished damage done by roe deer and other cervids? It’s probably impossible in case of young oaks maples and ashes.

Generally, during reading this paper, I did not learn too much about roe deer in Lithuania. Unfortunately, I feel that most of presented data is of limited scientific value.

no comments

Author Response

Reviewer 2 comments and answers

Comment: Figure 1. I understand that this map presents current borders of Lithuania, are data presented in figure 2 from 1934-1940  relevant to this area?

Answer: you are right, in the mentioned time period Lithuania had several parts of the country occupied, however, maps in scientific papers should not focus on political issues. Inclusion of changing political boundaries in maps can introduce bias and potentially distract from the ecological message of the paper. CAGR is not related to other variables than time and numbers, and we do not compare roe deer densities, therefore different area of the country has no importance.

To acknowledge your comment in full, we added note to the Figure 1 caption:

Note: Between 1934 and 1940, the territorial boundaries of the country were altered in the western and southeastern regions due to political factors.

Comment: I don’t understand how roe deer were counted. I found only information about snow-tracking and roe deer observations. Generally counting of free living animals is extremely difficult. Different methods give different results. It’s impossible to count animals precisely. Even using the most advanced, modern methods, the error of evaluation exceeds 25%. How is it  possible that population grew so quickly between 2016-2020? I am afraid that presented data may not have any scientific value, being probably only imagination of hunters. It’s not clear if it is data from the whole country or only from hunting grounds? The same problem is with data about hunting bag. It is official data from hunting reports but real harvest is unknown. Unregistered hunting and poaching is a common practice in the whole Europe. Hunting bag might be a density trend indicator, but only under some circumstances: the numbers of hunters and their hunting activity being stable during the whole analysed period, length of hunting season and methods being the same, shooting is unlimited by management plan.

Answer: I introduced some clarification. New text at Lines 115–121 looks:

Until 1990, the main survey method was snow-tracking, adjusted according to the number of animals recorded visually. Game counts in the snow by foresters and hunters were carried out in a coordinated manner, on the same day throughout the country, and the results obtained were aggregated at several levels. Subsequently, after 1990s, the population size was taken as the sum of the animals present in all hunting areas in the country, as reported by hunting clubs. After 2018, the snow-track index [38] is again used to analyse population trends, but not the number.

In addition to what we say in Line 122, “We understand that official figures may be biased, but this is the only source of game numbers. “, I added additional text in the end of review, a chapter on the Limitations of the study.

Comment: Also data about roadkill are questionable. Roe deer is relatively small and probably only small fraction of killed animals are officially reported.

Answer: Thank you for mentioning this, though actually roe deer is a big animal when it goes to roadkill, as the damage is significant. Nowadays such accident must be reported to get the insurance money, therefore not reported cases are not numerous. Previously, we did very broad survey of roadkills in the country (sampling effort over 300,000 km of driving by professional biologists), and found 56 roe deer carcasses being not registered by Traffic Service. Extrapolation show, how many accidents involving roe deer might actually be not registered by police. To clarify this, we added text after Line 246:

Extrapolating from unregistered roe deer roadkill numbers based on the sampling effort, it is estimated that approximately 9200 individuals were not recorded by the Service during 2002–2022 [67].

Comment: In Lines 345-349 there is information about roe deer pellet group counting, the biggest problem with this method is fact, that detection rate and accumulation period of pellet groups is different in different habitats. Because of this, its impossible to study roe deer habitat selection using this method.

Answer:

I acknowledge your comment, but as the manuscript is a review paper, it's essential to consider all relevant published materials to avoid bias in reference selection. To address your concern, I have added a chapter on the Limitations of the study at the end of the paper.

Regarding the issue of bias, I cannot fully agree with the notion that it's impossible to determine habitat preferences. Our study utilized a comprehensive approach, with total transect length exceeding 700 km across the country from west to east, conducted by four individuals simultaneously. To enhance pellet detection rates, we employed narrow transects (2 m wide for roe deer), conducted slow sampling, and worked in pairs. Additionally, we tested pellet decay rates in various habitats to ensure accuracy in our observations. We also included open habitats, not only forests, to the pellet count transects.

Comment: In section 3.9, how the forester distinguished damage done by roe deer and other cervids? It’s probably impossible in case of young oaks maples and ashes.

Answer:

To my best knowledge, foresters used several criteria to attribute damages to a specific deer species. They looked at tracks and defecations, the thickness of the browsed shoots (the most limited in the roe deer), the de-barking, and, finally, used knowledge of the number of deer in an area.

I'd like to inform you that during the period of 1960–1970, there were between 1100 and 6800 elk and 380–5300 red deer, with an uneven distribution across the country. Therefore, by employing a comprehensive approach, it was possible to attribute damage to specific species. I agree with you, biases were present, but I am just citing references.

After 1998, damage was no longer attributed to certain species of deer as numbers of moose and red deer increased, they inhabit all territory of the country.

To address your concern, I have added a chapter on the Limitations of the study at the end of the paper.

Comment: Generally, during reading this paper, I did not learn too much about roe deer in Lithuania. Unfortunately, I feel that most of presented data is of limited scientific value.

Answer: I made every effort to ensure that all references in several languages were included in this review paper, thereby reflecting the roe deer situation in the country based on all available knowledge about the species. As the only review of its kind in English, it will reach an international readership. I am pleased that other reviewers did not raise any concerns about its scientific value.

Reviewer 3 Report

The work is a review and focuses on many aspects of roe deer in Lithuania.
The paper aimed to describe roe deer in Lithuania, including the history of the species, its current status and abundance, ecology and other aspects of the species' management. The goal was achieved based on literature and statistical data on the volume of hunting harvest and the number of deer killed in road collisions.
In the chapter Results, the author presented historical data and current on the presence of roe in Lithuania, hunting data, roadkill in 2002–2022, environmental parameters and deer ecology, genetic diversity of Lithuanian deer, pathogens, damage to forest stands and roe deer population management.
The author has done much work and attempted to cover many aspects of the deer population. In this regard, I have no objections to the work. However, this topic is so vast that the author could not cover many points. Missing, for example, are ways of assessing individual conditions, differences concerning field and forest deer. This may be due to the lack of Lithuanian publications on roe deer.
I am not convinced that the work is suitable for publication in the journal Forests since this type of work is instead published in local professional journals. However, I leave the decision to the esteemed editors.

Author Response

Reviewer 3 comments and answers

Comment: The work is a review and focuses on many aspects of roe deer in Lithuania. The paper aimed to describe roe deer in Lithuania, including the history of the species, its current status and abundance, ecology and other aspects of the species' management. The goal was achieved based on literature and statistical data on the volume of hunting harvest and the number of deer killed in road collisions. In the chapter Results, the author presented historical data and current on the presence of roe in Lithuania, hunting data, roadkill in 2002–2022, environmental parameters and deer ecology, genetic diversity of Lithuanian deer, pathogens, damage to forest stands and roe deer population management.

The author has done much work and attempted to cover many aspects of the deer population. In this regard, I have no objections to the work. However, this topic is so vast that the author could not cover many points. Missing, for example, are ways of assessing individual conditions, differences concerning field and forest deer. This may be due to the lack of Lithuanian publications on roe deer.

Answer: Thank you for the positive evaluation of the manuscript. The topic you mention was not covered in the cited papers on the roe deer from Lithuania. To summarize drawbacks in the material used, I added chapter on limitations of the study in the end of article. Thank you for this comment, such chapter really adds some additional value the review.

Comment: I am not convinced that the work is suitable for publication in the journal Forests since this type of work is instead published in local professional journals. However, I leave the decision to the esteemed editors.

Answer: here, I fully agree with you, decisions are made by editors.

Reviewer 4 Report

The manuscript is written in great detail and I am mostly learning from the author rather than looking for problems. However, I think that this manuscript should be published in a regional journal, such as Lithuania's own, because the content of the manuscript is very targeted and the readers should be more within Lithuania.

Line 2 Doing a review of a particular animal in a country is necessary for researchers in that country, but publishing it in an SCI journal would have a very small readership, and it would be better to publish it in a journal in that country. In particular, there are Lithuanian words in the manuscript.

Lines 106-107 There are now many internet sources of information available for roadkill data. For example, some citizen science provides occasional roadkill encounters.

Lines 191-192 Are there some years where the data is missing? Are other data available for every year? Why can some places be connected with a line and others are disconnected?

Final question: the manuscript mentions a lot of things in great detail, which is good, but it seems hard to see where the focus is.

Minor editing of English language required

Author Response

Reviewer 4 comments and answers

Comment: The manuscript is written in great detail and I am mostly learning from the author rather than looking for problems. However, I think that this manuscript should be published in a regional journal, such as Lithuania's own, because the content of the manuscript is very targeted and the readers should be more within Lithuania.

Line 2 Doing a review of a particular animal in a country is necessary for researchers in that country, but publishing it in an SCI journal would have a very small readership, and it would be better to publish it in a journal in that country. In particular, there are Lithuanian words in the manuscript.

Answer: The Lithuanian words in the manuscript were made mandatory to better explain the key terms used in the literature search. As I understand it, this requirement applies to review articles based on published sources. This is the only way to achieve repeatability.

As regards publication in a regional journal, we agree with the editor's decision.

Comment: Lines 106-107 There are now many internet sources of information available for roadkill data. For example, some citizen science provides occasional roadkill encounters.

Answer: of course, I agree with you, but in Lithuania these sources are absent. On the other side, how many citizen scientists can see difference between young red deer, young moose and roe deer, when they are on the road after crash with heavy truck? Police data are at least partly confirmed by hunters, therefore more reliable.

Comment: Lines 191-192 Are there some years where the data is missing? Are other data available for every year? Why can some places be connected with a line and others are disconnected?

Answer: you are right, some data are missing or questionable (in 1943, who cared about wildlife survey in the middle of the war?). We adder text to the Figure 2 caption:

Data for 1940–1942 and 1944–1946 are not available, while data for 1943 are not reliable.

Comment: Final question: the manuscript mentions a lot of things in great detail, which is good, but it seems hard to see where the focus is.

Answer: in my opinion, which is confirmed by the information from many publishers, review papers can sometimes lack a focused aim for specific reasons (such as providing a comprehensive overview of a particular topic). This was my case, as I aimed to summarize all available information on a species in a country. Therefore, review paper encompassed a wide range of subtopics and research findings, synthesizing existing literature.

To explain this, I added some text to the “aim”, Line 62.

Round 2

Reviewer 2 Report

In the title there is an information that roe deer is the most abundant cervid, but in the paper there’s no data about other cervid species. In the case of study site, the data about area of Lithuania in different periods should be presented. Otherwise it is impossible to interpret data before 1945 and after the collapse of the soviet union.

I think that a researcher, while preparing a literature review paper, needs to assess the quality of available data and select them. From my point of view most of them is just imagination of hunters and forests. If they were obliged to complete document about game numbers or damages, they just did this, but from the scientific point of view it is not a reliable source of data. If politicians obliged hunter to count voles or frogs they would also do this precisely. It is highly interesting how hunters censused exact numbers of game animals on the basis of snow tracking? And how they combined this data with direct observation? It was of course (before climate change) good method for roe deer monitoring, but the results of this work is numbers of roe deer tracks/km/24h of snow cover. If there is data on this density estimator its worth being shown instead of figure 2. It is impossible to recalculate such data (i.e. an index of snow tracks) to density and abundance in case of roe deer. Maybe, during last 90 years, some scientific based procedure (driving census, FSC, FAR, REM, distance sampling, aerial census, DNA analysis, spot light counting…) were used for roe deer survey in Lithuania. Such data even in small scale are worth of being published or cited, especially as compared to official “data”. If hunters have counted tracks of roe deer for decades, there should be some original protocols filled during the fieldwork, it might be an interesting and valuable source of data. Such information should be statistically analysed. If hunters systematically observed roe deer on transects maybe it is possible to apply Distance Software to analysed this observation using distance sampling technique?  Such calculation would be reliable and raw field data should be available for researchers. If Lithuania is divided into permanent hunting grounds it’s easy to get some data about roe deer frequency of occurrence. For example in 1948 roe deer was present in 30% of analysed units, next in 60% and finally in 100% of them. It would be a valuable indicator of population development. If foresters in Lithuania discovered a key to distinguish which species browsed young oaks it also should be an object of scientific analyses by experts. It is not shown which forests does this data refer to – is this state forests, private forests etc.? Publishing such data would be informative, instead data which reliability is questionable. For me it is obvious that the Autor should work on raw data collected in the field by hunters and foresters, not on the final results reported by them.

Reference [38] link is unaffordable

Chapter 3.9 is really mysterious. Except the basic problem with determination which species harmed the tree, the next question is how foresters calculated damages. It is difficult to express it in hectares. In example, on 1 ha of young forest plantation there is 10000 oaks, and roe deer graze 2000 of them, in different places of this plantations. Did in such case forester report that 1 ha of plantation was destroyed by roe deer? They should count all grazed trees and calculate the percentage share of eaten trees. Which type of damage is reported by foresters? Was this a lethal damage, when trees have to be planted again, or just a moderate level of damage when trees can survive. Besides, level of damages is corelated with area of deciduous plantations established in a particular year. This data should be presented simultaneously.

On the basis of data about roe deer population productivity, numbers of roe deer, harvest, harvest rate, damage, it is obvious that such population cannot exist. Why damages was the highest when density was the lowest? If in Lithuania there is 175000 roe deer, density is extremely low, only around 2,7 individual/100 ha. I suppose that in reality the density is 5 or 10 times higher. Even if there is 175000 roe deer, presented level of harvest is very low. With such level of hunting bag, this population would grow about 25000 individuals yearly. We have to take under consideration that minimum error of roe deer census is 25%. In such case, level of harvest rate is included in the error of population abundance level. When showing the  numbers of individuals standard error or standard deviation should be presented. In the USA, among scientist, popular is acronym GIGO, i.e. garbage in garbage out. Simply, performing even the most advanced statistical calculation on unreliable data is senseless.

In line 212 there is information about decreasing of population between 1996-1997, but I cannot see this in the figure 2.

I am afraid that most of cited papers are prepared on the basis of questionable data. In example information about impact of wolf predation on roe deer population. It is very difficult and complex topic. We need detailed data about wolf and roe deer population, diet composition of wolf and many other data. I am afraid that quality data about wolf are similar to this about roe deer. It is an open question, if numbers of roe deer depend on numbers of predator or numbers of predators depend on number of roe deer? Also looking at relation between winter weather condition and roe deer survival is from scientific point of view complicated and requires large scale telemetry study.

Legitimisation of hunters imaginate data in a scientific, high quality journal, is simply harmful for game management and future research in Lithuania. If we realised that we do not have any data about abundance, population increase, sex and age structure, such knowledge might be a trigger to implement country wide system of game monitoring. Of course the only way is to track trends of population on the basis of relative index of abundance. Assessment of numbers of individuals is impossible on the country scale, which is obvious for anyone who even once tried to count ungulates. Even if we spend enormous amount of money, involve thousands of people and obtain some results, the error of such census will exceed 200-300%. Unfortunately, anachronic idea about game management from XIX century Germany, is still vital among hunters and politicians. It is impossible to implement procedure from livestock breeding to wildlife management, as it was intended to do, during last 200 years.

In the case of traffic accidents it is easy to imagine, that driver after collision, take the animal to the boot and uses it for consumption. Its might be a good compensation after collision with roe deer. In my country roe deer and hare are so-called “boot animals”. Also dogs and other predators can take away carcass. Roe deer can be lethally wounded but escape even several km from the road.

In oak-hornbeam forest it is easy to detect roe deer pellets, in the case of coniferous stands it is much more difficult because pellets simply “sink” in blueberry bushes. In alder forest detection depends on water level.

Y axis in figures are not described.           

In line 488 However, roe deer are not a commercial target in hunting grounds, what does it mean? Venison or trophies do not have any economic value? I found several offers for roe deer hunting in Lithuania, two day hunt cost 500-800 euro. https://www.bookyourhunt.com/en/roe-deer-hunting-in-lithuania. I also found offers for fallow deer hunting, if this species coexist with roe deer, pellet group identification is questionable.

I found publication about Lithuanian mammals (Hystrix, (n.s.) 8 (1-2) (1996): 9-15) There is crucial information about game monitoring “Official survey data may differ two-to-three times from actual populations in both directions (unpubl. survey data of Laboratory of Theriology, Institute of Ecology)” But, as a understand nothing changed during last three decades. Official survey are still “conducted” and publicly presented!

I really would like to read paper about roe deer in Lithuania, but it must be based on scientific data.

From my point of view the only reasonable way to revise this paper is to remove data about number of individuals and all grey literature, also questionable data about damages. If Author do this, the paper should be resubmitted and revised again. Proposal of scientific based, country scale, angulate monitoring program, also might be presented.

No comments