Isotopic Space of the House Mouse in the Gradient of Anthropogenic Habitats

Round 1

Reviewer 1 Report

Dear authors,

This is a very good attempt to study an interesting topic. Nevertheless, there are many things throughout the manuscript that should be improved in order to fully take advantage of the work you have performed.

The major problem is that samplings are very unbalanced and data are not properly representative of all the situations that are to be tested. Furthermore, there is a clear categorization of locations in terms of anthropization, but other sources of variation that appear to be important are not clearly stated, described, explained, or tested. Improvements are needed in the methods as well as the results sections, in order to make a stronger discussion of your results.  

Along the manuscript, you will find many annotations asking for clarification, more detail, or changes in the organization of contents, that could help improve the overall quality of the ms.

I hope the review and suggestions are useful for you.

Best regards,

 

Comments for author File: Comments.pdf

Author Response

Comment: This is a very good attempt to study an interesting topic. Nevertheless, there are many things throughout the manuscript that should be improved in order to fully take advantage of the work you have performed.

Answer: thank you, we incorporated all your suggestions, including those in pdf.

Comment: The major problem is that samplings are very unbalanced and data are not properly representative of all the situations that are to be tested. Furthermore, there is a clear categorization of locations in terms of anthropization, but other sources of variation that appear to be important are not clearly stated, described, explained, or tested. Improvements are needed in the methods as well as the results sections, in order to make a stronger discussion of your results.

Answer: unfortunately, balanced sampling was not possible. For example, house mice were not present in the orchards and plantations in 2019, 2020 and 2022. In one of the homesteads and in the kitchengarden house mice were trapped only in 2022, but not in 2018–2021. Therefore, for a balanced sample we could need next four years of trapping, but there is no guaranty we could trap more house mice. Therefore, we do our best to analyse data we have.

In the manuscript you reviewed we tried to cut all details, referencing details to our previous publications. As per your comments, details now are added to Methods, and explanations to Results.

Comment: Along the manuscript, you will find many annotations asking for clarification, more detail, or changes in the organization of contents, that could help improve the overall quality of the ms.

Answer: all comments were acknowledged, see revised text.

As for the comment about line 182 (after the Table 2), analysis was done on the pooled data, therefore age and gender structure did not depend from the insufficient sample size that much. The same is true in Lines 233 and 234, gender and age differences were analyzed on the pooled sample.

Comment at the Line 262: absence of intraspecific differences does not contradict to the statement about the niche width – it says that males, females and all age groups have equally wide trophic niche. Explanation added to the text.

As for the comment in Lines 290 to 297, we are discussing our own results, published in [52] and [57]. The bank vole is also omnivorous species, like house mouse, and was trapped in the same habitats, therefore, comparison of isotopic values is relevant. Our fault, we did not explain that we are comparing data of another species from the same small mammal communities. Some text added to clarify this.

Line 307 – apologies, in the version you reviewed we just pointed out, that „continuous predictors, body mass (T² = 0.001) and body condition (T² = 0.02), all were not significant” and did not include weak correlations to Results section previously, so now added. As r-squared is less than 10%, both factors are really not important.

Reviewer 2 Report

The manuscript is focused on the analysis of influence of individuals and ecological factor on ratio of N15 and C13 isotopes in house mouse hair. 

There are some ambiguities, which should be corrected or changed for improving comprehensibility of the manuscript.

Some habitats were analysed on pair of localities (some locality pairs were more than 100 km far) and each locality of the same habitat can have another vegetation cover and food supply. We cannot exclude also some influence of climate (north x south, distance from sea- and consequent differences in humidity, wind... etc.) on vegetation and food supply.  From these reasons I would recommend not to join localities of the same habitat together and analyse each locality separately. I realized, that higher number of individuals and smaller number of analysed places are better for finding statistical differences. However, numbers of individuals after joining are still too low for detailed analysis of habitat effect (with exception of category “farms” and perhaps all “orchards and plantations” together). Number of mice in other localities/habitats is too small (1 and 2 individuals for apple orchards, 4 individuals for meadow, 1 and 1 samples for currant plantation, 2 raspberry plantation, 1+4 homestead).  After analysis of each locality separately, there should be possibly discovered some difference between both farms, as the sample size of each locality is large enough and the distance between farms is 150 kms.

Lines 87-92- habitats are in order to increasing degree of anthropisation, where kitchen garden, homesteads, and farm are the most anthropogenic. However, there is not clear, where exactly traps were laid. In Line 160 is mentioned, that some habitats are indoor, but is not mentioned which ones. It is important to mention detailed information about trapping places (mainly farms, homestead, kitchen garden), as it is important for knowledge of potential food source of mouse (were traps laid to surrounding of houses?, into houses?, Stables- with cattle, poultry?, yards, buildings? herbs? corn?- for the whole year as food for farm animals or corn from the crop field-seasonal changes? Hay? vegetable? near forest? Meat?).

Lines 123-124: “Five per cent of samples were analysed duplicated”- as there were 105 samples together, I feel, that more illustrative should be: “five samples out of 105 total”

Line 141: there is not clear, what exactly means variables: “food origin” and “season” (year or date)? Date should be included into statistical model, as season determine availability of food supply (such as vegetation cover, seeds, etc…)

Table 2: Superscript letters should indicate differences, however it is unclear between which pairs of habitats, as there are five times “a” and two times “b” in column P%. How can be calculated statistical differences between Shannon’s index (when there is one number for each locality)???

 

Lines 180-182- there is no surprise, that ranges for homestead and kitchen gardens were narrow in comparison with farms, as in farms were captured 86 individuals, however in homestead five and kitchen garden three individuals. In discussion should be detailed explanation, if trapping on farms were on more divergent microhabitats or if there were uniform places (stable, house, surrounded orchard, garden, barn, rick of straw?).

Author Response

Comment: Some habitats were analysed on pair of localities (some locality pairs were more than 100 km far) and each locality of the same habitat can have another vegetation cover and food supply. We cannot exclude also some influence of climate (north x south, distance from sea- and consequent differences in humidity, wind... etc.) on vegetation and food supply. From these reasons I would recommend not to join localities of the same habitat together and analyse each locality separately. I realized, that higher number of individuals and smaller number of analysed places are better for finding statistical differences. However, numbers of individuals after joining are still too low for detailed analysis of habitat effect (with exception of category “farms” and perhaps all “orchards and plantations” together). Number of mice in other localities/habitats is too small (1 and 2 individuals for apple orchards, 4 individuals for meadow, 1 and 1 samples for currant plantation, 2 raspberry plantation, 1+4 homestead).  After analysis of each locality separately, there should be possibly discovered some difference between both farms, as the sample size of each locality is large enough and the distance between farms is 150 kms.

Answer: sample size is not “too small” – just mice were not present all the time and in all habitats. In the kitchengarden and in one of the two homesteads house mice were trapped only in 2022; house mice were present in the orchards and plantations in 2018 and 2021, but not present in 2019, 2020 and 2022. We could sample next four years, but there is no guaranty we could trap more house mice. Small paragraph added before Table 2.

We of course did ANOVA with sampling site as grouping factor, and post-hoc analysis, with none of the significant differences. In one of our papers we analysed small mammal communities in Lithuania with geographic aspect. There are habitat-based differences in relative abundances and proportions of dominant species, which are much stronger than location-based ones. The same is valid for stable isotopes – their values depend on the factors we present in the paper, but we omit those factors which were not significant.

In short – farm is farm, and sources of additional food were similar in both farms we analyse, so there were no significant differences in the mice diet. Therefore we can only extend the text, explaining there were no differences, but cannot add more results to the manuscript.

As house mice are very commensal (there are no permanent feral populations of house mice in Lithuania), this mostly eliminates influence of the climate.

As for the food supply you mention in the comment, we added text explaining more details.

Comment: Lines 87-92- habitats are in order to increasing degree of anthropisation, where kitchen garden, homesteads, and farm are the most anthropogenic. However, there is not clear, where exactly traps were laid. In Line 160 is mentioned, that some habitats are indoor, but is not mentioned which ones. It is important to mention detailed information about trapping places (mainly farms, homestead, kitchen garden), as it is important for knowledge of potential food source of mouse (were traps laid to surrounding of houses?, into houses?, Stables- with cattle, poultry?, yards, buildings? herbs? corn?- for the whole year as food for farm animals or corn from the crop field-seasonal changes? Hay? vegetable? near forest? Meat?).

Answer: our apologies for oversimplified presentation of methods. We extended presentation of Material and Methods. In short – mice were trapped in summer and autumn seasons in commercial orchards and plantations (as far as we know from results of over 40-year-long investigations of small mammals in the country, feral mice do not exist in Lithuania in winter). In the commensal habitats mice were trapped in all four seasons. We also add text explaining sources and availability of additional food sources in these commensal habitats.

Comment: Lines 123-124: “Five per cent of samples were analysed duplicated”- as there were 105 samples together, I feel, that more illustrative should be: “five samples out of 105 total”

Answer: we changed text as suggested.

Comment: Line 141: there is not clear, what exactly means variables: “food origin” and “season” (year or date)? Date should be included into statistical model, as season determine availability of food supply (such as vegetation cover, seeds, etc…)

Answer: there are three ways of mice food origin – natural foods, human foods and livestock feed. We rewrote text to explain details.

Seasons were spring, summer, autumn and winter. There is no advantage to include date, as in agricultural habitats trapping was done only twice er year, in summer and in autumn. Therefore, we use season, not the date of trapping, also for commensal habitats.

Comment: Table 2: Superscript letters should indicate differences, however it is unclear between which pairs of habitats, as there are five times “a” and two times “b” in column P%. How can be calculated statistical differences between Shannon’s index (when there is one number for each locality)???

Answer: in the Table 2 different superscript letters denote significance of difference column-wise, as a standard approach. Thus, in column P% there are five small values, denoted by “a”, and two large values denoted by “b”. All “a” did not differ, all “b” are the same, but all “a” differ significantly from all “b”.

Calculation of significance of differences between Shanon’s index is standard, it includes procedure of bootstraping, presented in PAST software. We added text to explain this in detail.

Comment: Lines 180-182- there is no surprise, that ranges for homestead and kitchen gardens were narrow in comparison with farms, as in farms were captured 86 individuals, however in homestead five and kitchen garden three individuals. In discussion should be detailed explanation, if trapping on farms were on more divergent microhabitats or if there were uniform places (stable, house, surrounded orchard, garden, barn, rick of straw?).

Answer: in fact, homesteads and farms differed mainly in the form of available additional food: human food in homesteads, human food and livestock feed in farms. In both farms and homesteads, natural foods were available.

Therefore, we see that not the analyzed mice numbers, but the variety of foods was responsible for the variance mentioned in lines 180–182. If you look further, at Figure 3, central niches depend on the food source, and central niche related to availability of the livestock feed is the narrowest one.

We agree with your comment, and add explanations to the text in Results, not only Discussion. Thank you for pointing out incompleteness of our text.

Round 2

Reviewer 2 Report

The manuscript is sufficiently clear, all my comments have been reflected, I have no other comments.