The Value of Urban Nature in Terms of Providing Ecosystem Services Related to Health and Well-Being: An Empirical Comparative Pilot Study of Cities in Germany and the Czech Republic

Round 1

Reviewer 1 Report

Overall:

This study investigates subjective perceptions and evaluation of citizens to urban green spaces in terms of their well-being. The motivation of interviewees to visit green space and time spent in green spaces was identified using surveys in the three cities in Germany and Czechia.

The strength of the study is a large number of types of urban green space and surveys carried out in the two countries. However, the main limitations of this study that were not mentioned are that the data on visiting time, proximity from residence to green were self-reported, there is limited information about visitors’ socioeconomic characteristics, health, and other individual information.

The relevance and originality of this study should be highlighted, it should be highlighted how this study differs from similar ones already conducted, what is the novelty, how it contributes to the already existing knowledge.

Material and methods:

In the materials and methods section, when discussing the population in different cities, different years are cited (2019), the information of inhabitants for all three cities should be for the same year.

The description of urban green areas where surveys were carried out should be expanded with a more detailed overview of their size, biodiversity, its composition, predominant plant species, facilities, and other characteristics as the size of urban green space is important for ecosystem services.

There is also a lack of information on how survey areas of green spaces in different cities were selected, as well as why exactly such 14 landscape element types were analysed.

Results:

It would be useful to present descriptive statistics on demographic factors and other individual characteristics of study participants in different cities.

One of the hypotheses of this study was to assess if spending time or walking in green spaces promotes positive thinking and helps reduce stress, however, there is a lack of statistical analysis to assess how the well-being of visitors depends on the time spent in urban green spaces.

It would also be interesting to determine the influence of the visiting time and the distance from the residence place to the green space on the perception of visitors, their personal well-being related to different landscape elements and the main reasons for visiting green space.

There is evidence that the same area of green spaces near the centre provides much higher cultural services than that near the outskirts of the city. It would be important to analyse how the most important ecosystem services differed depending on the distance of the surveyed green areas to the centre of the city.

In addition to one of the drawbacks of the limited data collected on respondents’ demographic and socio-economic factors, it is important to assess the main differences in terms of the value and perception of green spaces between the gender and education level.

Discussion:

The authors state that the duration of time spent in green space is related to the size of the site. It is important to present the characteristics of the analysed green areas, describing their area, assessing their main functions, condition, and other characteristics.

The authors state that “activities such as dog walking were more often named as unimportant than important”, I wonder if this conclusion was made by interviewing and evaluating only those who indicated they had a dog or all the respondents and what proportion of them had a dog?

I see much more study limitations than authors mentioned: the data on visiting time, proximity from residence to green were self-reported, there is limited information about visitors’ socioeconomic characteristics, health, and other individual information.

Minor changes:

Introduction:

Page 2, the end of the third paragraph: “…, see van den [21].”.

Author Response

Reviewer 1

Overall:

This study investigates subjective perceptions and evaluation of citizens to urban green spaces in terms of their well-being. The motivation of interviewees to visit green space and time spent in green spaces was identified using surveys in the three cities in Germany and Czechia.

The strength of the study is a large number of types of urban green space and surveys carried out in the two countries. However, the main limitations of this study that were not mentioned are that the data on visiting time, proximity from residence to green were self-reported, there is limited information about visitors’ socioeconomic characteristics, health, and other individual information.

Answer: Thanks for the appreciation of our work. Using the self-reported visiting time and proximity seems to be necessary, because an objective measurement cannot be done without a violation of data privacy. If we wanted to investigate the self-reported health status, we would need an additional set of questions in the questionnaire. This would make our questionnaire much longer and the willingness to participate accordingly lower. Since we also wanted to address ecosystem services, we had to skip that issue.

The available information on socio-economic characteristics and other individual information have been added (Tab. 1, line 338)

The relevance and originality of this study should be highlighted, it should be highlighted how this study differs from similar ones already conducted, what is the novelty, how it contributes to the already existing knowledge.

Answer: A sentence was added to highlight the originality (line 139-141)

Material and methods:

In the materials and methods section, when discussing the population in different cities, different years are cited (2019), the information of inhabitants for all three cities should be for the same year.

Answer: It is however depending on reported from either December 31^st or January 1^st, the manner differs between Czechia and Germany, so it must not be the same year; we unified the data following official statistics (line 194-203).

The description of urban green areas where surveys were carried out should be expanded with amore detailed overview of their size, biodiversity, its composition, predominant plant species, facilities, and other characteristics as the size of urban green space is important for ecosystem services.

Answer: An extended description of the urban green areas where surveys were carried out have been added (Now: Annex 1)

There is also a lack of information on how survey areas of green spaces in different cities were selected, as well as why exactly such 14 landscape element types were analysed.

Answer: Explanations how we selected survey areas and landscape element types have been added (line 182-188, 306-307)

Results:

It would be useful to present descriptive statistics on demographic factors and other individual characteristics of study participants in different cities.

Answer: The available information on socio-economic characteristics and other individual information have been added (Tab. 1, line 338)

One of the hypotheses of this study was to assess if spending time or walking in green spaces promotes positive thinking and helps reduce stress, however, there is a lack of statistical analysis to assess how the well-being of visitors depends on the time spent in urban green spaces.

Answer: Hypotheses have been removed and replaced by research questions. However, several statistical analyses and tests have been added where meaningful (line 322-332, 358-378, 394-422, 446-458, 601-620).

It would also be interesting to determine the influence of the visiting time and the distance from the residence place to the green space on the perception of visitors, their personal well-being related to different landscape elements and the main reasons for visiting green space.

Answer: That was actually one of the research questions of the study, but this topic is rather complex, so there are different answers. The paper provides answers to different aspects such as the distances (line 394-403), the personal well-being (Tab. 2 and 3), or the landscape elements (Tab. 7, Fig. 10).

There is evidence that the same area of green spaces near the centre provides much higher cultural services than that near the outskirts of the city. It would be important to analyse how the most important ecosystem services differed depending on the distance of the surveyed green areas to the centre of the city.

Answer: The small number of survey areas, partly in almost similar distance to the city centres and their different characteristics (park with and without blue elements, meadows with and without river, forests) make it difficult to answer this question. Nevertheless, we calculated the distance from city centre (as place of town hall) (Annex 1) and the connection to the ecosystem services. So, we found only one slight connection for Dresden, which is larger than the other two cities where the effect, if it exists, is not reflected in the few locations under study (line 608-620).

In addition to one of the drawbacks of the limited data collected on respondents’ demographic and socio-economic factors, it is important to assess the main differences in terms of the value and perception of green spaces between the gender and education level.

Answer: logit models that predict the reported feelings based on socio-demographic characteristics (age, gender, education, income) have been added to the paper (see Table 3 and description of results line 358-378).

Discussion:

The authors state that the duration of time spent in green space is related to the size of the site. It is important to present the characteristics of the analysed green areas, describing their area, assessing their main functions, condition, and other characteristics.

Answer: An extended description of the urban green areas where surveys were carried out have been added (Annex 1).

The authors state that “activities such as dog walking were more often named as unimportant than important”, I wonder if this conclusion was made by interviewing and evaluating only those who indicated they had a dog or all the respondents and what proportion of them had a dog?

Answer: Thanks for this good question! We added this information how many people had dogs (Tab. 1, line 338) and to the relation (lines 601-607).

I see much more study limitations than authors mentioned: the data on visiting time, proximity from residence to green were self-reported, there is limited information about visitors’ socioeconomic characteristics, health, and other individual information.

Answer: The limitations have been mentioned (lines 726-728). The available information on socio-economic characteristics and other individual information have been added (Tab. 1, line 338).

Minor changes:

Introduction:

Page 2, the end of the third paragraph: “…, see van den [21].”.

Answer: Mistake has been deleted.

Reviewer 2 Report

The research problem that is the subject of the work is interesting and probably necessary. However, the work in its present form cannot be published. The authors studied three cities and is insufficient to generalize the results on the populations. The main objection to the work is the complete lack of statistical quantitative methods. The work may be printed if it concerns information about the preferences of people in three cities.

Please find below detailed comments:

1. The work should clearly define its purpose. Perhaps the most relevant to the topic is the purpose given in the executive summary [line 26]. but it was not achieved. The other aims  [line 41], [line 122]. What was the purpose of the work?
2. Lines 134-147 should be included in the methodology of work. The selection of 3 cities for the research, in my opinion, does not allow for generalizing the conclusions and comparing countries.
3. The selection of the target group included only people who used green areas [lines 168-176]. I am not convinced by these conclusions. For example, if you did the tests in a hospital, you would find out that most people are so sick. So in my opinion, the entire study was wrongly founded. In addition, the entire study covers only three cities that were selected biased. It seems to me that you can only speak and judge people from these cities. In addition, the study should also cover people who do not use urban green areas.
4. Placing figure 2 on page 9 makes it very difficult to accept the work. The authors refer to the figure on page 5 [line 181]. Likewise, Figure 3 [line 199] and 4 [line 205]
5. Will the Annex be attached to the publication? [line 217]
6. Where is table 7? [line 249]
7. In my opinion, there is no difference [line 272]. It cannot be argued that the results are similar in other cities [line 270]
8. The value of 55% "more energetic" is not referenced in Table 2 [line 269]
9. This part is unclear to me. Where are the Linkert scores included? [lines 280-285] What is something else in Table 2?
10. I do not fully agree with the conclusions regarding Table 2. The worst rated place is Park u Machovek (authors quote Přehrada Harcov), but I am not sure if the result can be considered reliable due to the small number of respondents (22).
11. Why these strange results [lines 317-318]. After calculating my own, I got 11:10 and 9:39
12. This part of the work should probably be earlier [lines 339-343]
13. I don't understand this part. Where did these percentages come from and how do they compare to Table 5 and Figure 7. [lines 410-421]. Why is the figure only showing the results for Dresden? What do the abbreviations RI, RIN, NRI, NR, NI mean.
14. Has it been shown by the authors? [line 438-411] Has this been demonstrated in other works?
15. 
    This part of the work confirms that the experiment was methodologically incorrectly carried out. Contradictory results are obtained here [lines 453-471], which the Authors are trying to explain at all costs. This has nothing to do with the second hypothesis.
16. What is this paragraph about? [line 507-514]
17. This part of the work should be in the introduction [lines 516-531]
18. The authors try to generalize their research. In this case, we can only talk about the situation in two Czech cities and Dresden. A methodological error is the generalization of these studies on a larger scale.
19. The first sentence has not been proven, these are subjective assessments of the respondents [lines 543-545]
20. On what basis did the authors draw their conclusions [lines 561-596]
     They have a very loose connection with the research carried out.

Author Response

Reviewer 2

Comments and Suggestions for Authors

The research problem that is the subject of the work is interesting and probably necessary. However, the work in its present form cannot be published. The authors studied three cities and is insufficient to generalize the results on the populations. The main objection to the work is the complete lack of statistical quantitative methods. The work may be printed if it concerns information about the preferences of people in three cities.

Answer: We used statistical methods and added statistical information. It was not possible in the frame of the given project to provide surveys that can claim representativity for the whole city population. We only addressed people that were within the urban green areas (not all the others see lines 181-182). We did not try to generalize the results on the entire populations, in order to avoid such misunderstanding, the conclusions have been reformulated everywhere needed (lines 742-744, 790, 800, 805-806).

Please find below detailed comments:

1. The work should clearly define its purpose. Perhaps the most relevant to the topic is the purpose given in the executive summary [line 26]. but it was not achieved. The other aims  [line 41], [line 122]. What was the purpose of the work?

Answer: The purpose of work (line 136-139) was refined and replicated to the abstract (lines 35-37), the indirect aim was formulated (hopefully) clearer as “This article shall contribute to recognizing importance of specific green spaces as a factor of cultural urban development, and make recommendations that will benefit public health and the quality of urban living” (lines 51-54).

2. Lines 134-147 should be included in the methodology of work. The selection of 3 cities for the research, in my opinion, does not allow for generalizing the conclusions and comparing countries.

Answer: The according lines had to be deleted due to the request of other reviewers. Instead of hypotheses, we formulated additional research questions refining the purpose of work. We did not try to generalize our findings to two countries, sorry for causing this misunderstanding! The sentence about the two countries was deleted.

3. The selection of the target group included only people who used green areas [lines 168-176]. I am not convinced by these conclusions. For example, if you did the tests in a hospital, you would find out that most people are so sick. So in my opinion, the entire study was wrongly founded. In addition, the entire study covers only three cities that were selected biased. It seems to me that you can only speak and judge people from these cities. In addition, the study should also cover people who do not use urban green areas.

Answer: The study was directed to people currently using green areas only. We mad this clearer (lines 181-182). When we asked people how they feel right now and others how the think they would feel if they were there, the answers cannot be comparable since it would be a comparison between feelings in fact and assumptions. We consider it not meaningful to ask people not using green areas how they feel there, and how green areas may affect them because the answers would be biased by remnants and guesswork. We did not try to generalize the results on the entire populations, in order to avoid such misunderstanding, the conclusions have been reformulated everywhere needed (lines 742-744, 790, 800, 805-806).

On contrary to your claim the results would actually be biased if the study included people who do not use urban green areas. How are their opinions relevant for e.g., planning if they do not go there in the first place? How is your example with a hospital relevant? A much better example would be let’s say a bakery that wants to know people’s opinion on their cakes in order to increase profits. How would it help them to run a survey among people who do not buy cakes? You can argue that they might try to win new customers, but these should always be two separate surveys.

4. Placing figure 2 on page 9 makes it very difficult to accept the work. The authors refer to the figure on page 5 [line 181]. Likewise, Figure 3 [line 199] and 4 [line 205]

Answer: Good idea, the former figures 1 by 4 (new numbers 1, 3, 5, 7) were shifted upwards.

5. Will the Annex be attached to the publication? [line 217]

Answer: Yes, the Annex (Now: Appendix B) is part of the publication.

6. Where is table 7? [line 249]

Answer: We corrected the spelling mistake; it was Tab. 6; now it is Tab. 8.

7. In my opinion, there is no difference [line 272]. It cannot be argued that the results are similar in other cities [line 270]

Answer: We deleted the sentence.

8. The value of 55% "more energetic" is not referenced in Table 2 [line 269]

Answer: We balanced the rounding difference.

9. This part is unclear to me. Where are the Linkert scores included? [lines 280-285] What is something else in Table 2?

Answer: The Likert scores were included in the former Tab. 2 (now Tab. 4) column “Positive feelings reported (3=max.)”, the other data in Tab. 2 are the Number of respondents in each area, the percentage of people that see the respective green area as final destina­tion, the others were on the way “from A to B” (given as “Pathway”), the satisfaction mark for the respective green area and the calculated visiting time using the frequency and the duration of stay per month (lines 384-391).

10. I do not fully agree with the conclusions regarding Table 2. The worst rated place is Park u Machovek (authors quote Přehrada Harcov), but I am not sure if the result can be considered reliable due to the small number of respondents (22).

Answer: Přehrada Harcov is the green space with last satisfaction in Liberec, Park u Máchovky belongs to Děčín and is not the highest ranked there. But we added Park u Máchovky as well (line 432).

11. Why these strange results [lines 317-318]. After calculating my own, I got 11:10 and 9:39

Answer: Yes, the calculation is difficult, one must regard the different numbers of respondents in the several areas. Indeed, it is 12:10 and 8:59, we corrected the first number accordingly. 11:06 would be the average of all three cities.

No	Time	Sume time	Average
37	07:42:00	284:54:00
38	08:06:00	307:48:00
63	09:13:00	580:39:00
53	06:34:00	348:02:00
24	16:15:00	390:00:00
20	10:09:00	203:00:00	08:59
153	11:38:00	1779:54:00
21	14:16:00	299:36:00
27	04:29:00	121:03:00
54	17:20:00	936:00:00
18	07:36:00	136:48:00
22	08:03:00	177:06:00
80	12:29:00	998:40:00
84	13:35:00	1141:00:00	12:10
694	147:25:00	7704:30:00	11:06:06

12.This part of the work should probably be earlier [lines 339-343]

Answer: Sorry, here seems to be a misunderstanding. This answer is not to the question for the appreciation of green areas where we met people, but we asked for that (also other) green space in the city the respondent prefer the most. Therefore, the place might be correct here. We added a sentence to make it clearer (line 490-491).

13. I don't understand this part. Where did these percentages come from and how do they compare to Table 5 and Figure 7. [lines 410-421]. Why is the figure only showing the results for Dresden? What do the abbreviations RI, RIN, NRI, NR, NI mean.

Answer: We added a paragraph that explains the values in former Tab. 5 (new: Tab. 7) and the types in Fig. 7 (new: Fig. 10) (lines 565-576). Since we do not have the same basic data as for Dresden (the German LBM-DE) also for Czech cities, we decided to omit maps of the landscape elements and their effects (such as Fig. 9 and 10) there, since that would not be possible without a loss of quality.

14. Has it been shown by the authors? [line 438-411] Has this been demonstrated in other works?

Answer: This was shown by the cited literature, we added the clause: “As we know from literature [12, 46], …” (line 627).

15. This part of the work confirms that the experiment was methodologically incorrectly carried out. Contradictory results are obtained here [lines 453-471], which the Authors are trying to explain at all costs. This has nothing to do with the second hypothesis.

Answer: Unexpected results should not prove that the methodical concept is wrong, on the contrary. We are convinced that our argumentation is logical here and regret that the reviewer sees contradictions. Since we asked more than hundred people (at least around Přehrada Harcov) it is though possible that some visitors express critics although others prefer the same location.

16. What is this paragraph about? [line 507-514]

Answer: This paragraph is about the future of allotment gardens, which is a huge issue in Germany since the meanings in society are really contradicting. We collect arguments to suggest urban planners and city governments to keep gardens within the city but change a part of them to community gardens such as Golgi park to ease the above-mentioned conflict. The discussion part is used to derivate recommendations (in conclusions) from the results.

17. This part of the work should be in the introduction [lines 516-531]

Answer: The first sentence was shifted to introduction (lines 73-75), the rest should remain in discussion since it deals with our results.

18. The authors try to generalize their research. In this case, we can only talk about the situation in two Czech cities and Dresden. A methodological error is the generalization of these studies on a larger scale.

We did not want to generalize the results on the entire populations, in order to avoid such misunderstanding, the conclusions have been reformulated everywhere needed (lines 742-744, 790, 800, 805-806).

19. The first sentence has not been proven, these are subjective assessments of the respondents [lines 543-545]

Answer: The first sentence says with the clause “for self-reported well-being and mental health” that it is exactly subjective assessment which can only be proven by the answers of respondents. We added a clause constraining the meaning to the topic our research was focused on (line 741).

20. On what basis did the authors draw their conclusions [lines 561-596]
 They have a very loose connection with the research carried out.

Answer: The conclusions are drawn by the results and discussion. They deal only with the issues  of the survey we carried out, see Appendix B and sections 3 by 4 such as: proximity of the green space (question 1), length of stay (question 4), frequency of visit (question 3),  purpose of visit (question 6), accessibility of the area (question 2), most popular types of green spaces (question 10), complaints on shortcomings (question 5), connections to socio-demographical characteristics such as age and family (question 16 and 19), valuation of ecosystem services (question 11).

Reviewer 3 Report

The subject of the research is the impact of green areas in three selected cities on the health and well-being of people (by assumption of the inhabitants of these cities or the immediate vicinity, although it is not precisely defined). The survey method was used to collect statistical data. In modern times, the discussed research problem is very important, which was quite well justified in the article. However, in its present form, the article is not suitable for publication due to many weaknesses, mainly methodological. The following points require clarification:

1. The purpose of a scientific article cannot be to promote something (lines 41-43), even if it is to promote the noblest thing in the world. The text should present the results of scientific research, the methodology used, assumptions made, etc. Promotion of research results is another issue.
2. The purpose of the research has not been precisely explained. Various research objectives are given in several places: the aforementioned promotion, issues of spatial planning of green areas, evaluation of urban greenery by residents (lines 118-123), the importance of ecosystem services (428-429). There are no references to such broadly defined goals in the research results, e.g. can the study provide detailed guidelines for planning green spaces in specific locations of the cities under consideration? Such questions are not included in the questionnaire.
3. While the research hypotheses have been quite clearly defined, they have not been statistically verified, and this calls into question most of the discoveries made by the Authors. No statistical tests have been performed.
4. The weakest point of the article is the research methodology. Virtually no statistical research has been carried out. Calculating percentages, summing up, and averaging are not statistical studies. What research method was used? Why was the data entered into the statistical package?
5. The statistical population in the three examined cities was not precisely defined. The authors mistakenly call it the target group (line 177). Are they residents of these cities or also people from outside the region who temporarily use green areas? Or maybe also foreign and domestic tourists? On the basis of what criteria were people divided into random and on purpose (line 221)? The survey questionnaire does not specify this. If such data are missing, it means the population is undefined and the focus should be on sample representativeness, which I will discuss in the section below.
6. The research uses three statistical samples, one for each city. Unfortunately, nothing is known about the representativeness of these samples, however it is crucial for the reliability of the results. The most universal statistical formula, used to establish the representative sample size of the questionnaire survey, has the following form:

N(min) >= Z^2(α)/4E^2

where: N(min) – minimal sample size, Z(α) – statistical value of the Normal Distribution table, 1 – α confidence level, E – margin of error (the confidence interval). In this case, it is assumed that the information regarding the estimated structure indicator is missing. The starting point for the research is to demonstrate the representativeness of three samples. Otherwise, further calculations are pointless. For commonly accepted values, 1 – α = 0.95 and E = 0.05, the condition of the minimum sample size is not met in three examined cases:

N(min) >= (1.96)^2/4(0.05)^2 = 384.16 (385) < (235, 255, 204)

The sample size calculator can also be used to calculate the minimum sample size (https://www.surveysystem.com/sscalc.htm#two). Also, you can use a calculator to see what the margin of error will be for each sample. Taking into account the sample for Dresden, i.e. N = 235, we get 6.39%, which is above the standard (5%). This situation must be thoroughly explained. In special cases, where it is not possible to increase the sample size, such assumptions are acceptable.

7. What is the sampling process? It is necessary to explain that the sample is random.
8. The Authors indicate that the rest of the research results will be published in the following articles. It is important to select appropriate statistical methods in such a case. So far, all of the variables studied are categorical, so they should be placed on the nominal scale in the Stevens’ classification system (Stevens, S.S. On the theory of scales of measurement. Science 1946, 103, 677–680; Stevens, S.S. Psychophysics: Introduction to Its Perceptual, Neural, and Social Prospects; Transaction Publishers: New Brunswick, NJ, USA, 2008; ISBN 978-0-88738-643-5). In this case, the best research methods are the Pearson’s Chi-square test of independence and correspondence analysis.

The article is quite neatly written, contains many interesting details and precise maps of the studied cities with the indication of recreational areas, therefore it brings some added value, although not too great, to the literature on the subject. The only chance to save the text is to treat it as preliminary, pilot research, which should be clearly stated. After clarifying the doubts detailed in the review, there remains the question of the representativeness of the three samples. Using the sample size calculator (https://www.surveysystem.com/sscalc.htm#one) allows you to determine the margin of error (the confidence interval) for each sample with the numbers given in the article. For samples consisting of 235, 255 and 204 units, these margins, for 1 - α = 0.95, are respectively 6.39%, 6.14% and 6.86% (at percentage = 50%). This exceeds the standard values adopted in this type of research (5%), so I leave the decision to publish such an article to the Editor.

Author Response

Reviewer 3

The subject of the research is the impact of green areas in three selected cities on the health and well-being of people (by assumption of the inhabitants of these cities or the immediate vicinity, although it is not precisely defined). The survey method was used to collect statistical data. In modern times, the discussed research problem is very important, which was quite well justified in the article. However, in its present form, the article is not suitable for publication due to many weaknesses, mainly methodological. The following points require clarification:

1. The purpose of a scientific article cannot be to promote something (lines 41-43), even if it is to promote the noblest thing in the world. The text should present the results of scientific research, the methodology used, assumptions made, etc. Promotion of research results is another issue.

Answer: Thank you for the clarification! The purpose of work (line 136-138) was refined and duplicated to the abstract (lines 35-37). Actually, we do want that our research results can be useful for something beyond gaining knowledge. Isn’t that the purpose of each applied science?

2. The purpose of the research has not been precisely explained. Various research objectives are given in several places: the aforementioned promotion, issues of spatial planning of green areas, evaluation of urban greenery by residents (lines 118-123), the importance of ecosystem services (428-429). There are no references to such broadly defined goals in the research results, e.g. can the study provide detailed guidelines for planning green spaces in specific locations of the cities under consideration? Such questions are not included in the questionnaire.

Answer: The purpose of work (line 136-138) was refined and duplicated to the abstract (lines 35-37), the indirect aim (line 51) was formulated (hopefully) clearer as “the article shall contribute to…”. Some contributions are possible and were given in the result section.

3. While the research hypotheses have been quite clearly defined, they have not been statistically verified, and this calls into question most of the discoveries made by the Authors. No statistical tests have been performed.

Answer: Hypotheses have been removed and replaced by research questions. However, several statistical analyses and tests have been added where meaningful (line 322-332, 358-378, 394-422, 446-458, 601-620).

4. The weakest point of the article is the research methodology. Virtually no statistical research has been carried out. Calculating percentages, summing up, and averaging are not statistical studies. What research method was used? Why was the data entered into the statistical package?

Answer: We enhanced the description of research methodology (section 2). We did some further statistics were meaningful (line 322-332, 358-378, 394-422, 446-458, 601-620). Partly, we value percentages, averages and frequencies as reasonable and instructive descriptions of a survey outcome and didn’t want to delete them all. We run several logit models to identify possible relationships between reported feelings and socio-demographic characteristics. For most parts however, it makes little sense to go beyond descriptive statistics and correlations.

5. The statistical population in the three examined cities was not precisely defined. The authors mistakenly call it the target group (line 177). Are they residents of these cities or also people from outside the region who temporarily use green areas? Or maybe also foreign and domestic tourists? On the basis of what criteria were people divided into random and on purpose (line 221)? The survey questionnaire does not specify this. If such data are missing, it means the population is undefined and the focus should be on sample representativeness, which I will discuss in the section below.

Answer: The target group is not all citizens of the three cities, rather than the study was directed to people currently using green areas only. We added this clause (line 181-182) for clarification.

6. The research uses three statistical samples, one for each city. Unfortunately, nothing is known about the representativeness of these samples, however it is crucial for the reliability of the results. The most universal statistical formula, used to establish the representative sample size of the questionnaire survey, has the following form:

N(min) >= Z^2(α)/4E^2

where: N(min) – minimal sample size, Z(α) – statistical value of the Normal Distribution table, 1 – α confidence level, E – margin of error (the confidence interval). In this case, it is assumed that the information regarding the estimated structure indicator is missing. The starting point for the research is to demonstrate the representativeness of three samples. Otherwise, further calculations are pointless. For commonly accepted values, 1 – α = 0.95 and E = 0.05, the condition of the minimum sample size is not met in three examined cases:

N(min) >= (1.96)^2/4(0.05)^2 = 384.16 (385) < (235, 255, 204)

The sample size calculator can also be used to calculate the minimum sample size (https://www.surveysystem.com/sscalc.htm#two). Also, you can use a calculator to see what the margin of error will be for each sample. Taking into account the sample for Dresden, i.e. N = 235, we get 6.39%, which is above the standard (5%). This situation must be thoroughly explained. In special cases, where it is not possible to increase the sample size, such assumptions are acceptable.

Answer: We agree, that the sample size is lower than the minimum sample size you mentioned. The sample size is smaller due several reasons: (i) we applied the Simple random sample. The number of visitors of selected green areas was too low to meet the minimum sample size without asking of each visitor. (ii) The low number of people outside was partly caused by the hot temperature in time of data collection (especially in August 2018, when we carried out the main data collection). We didn’t expect these two situations before running the survey. Unfortunately, due the budget and staff limitation we were not able to meet the minimum sample size. We were focusing to cover all possible days and times in each city to avoid elimination of any group of possible respondents/visitors of parks, therefore in each city we spent more than one week with active data collection. This led to situation we covered all possible resident´s groups. Based on basic comparison of structure of residents in all three cities, the sample size is very similar to the structure in each city (when children are not involved), the deviations are mainly due to focus on people, currently using the urban green areas.

7. What is the sampling process? It is necessary to explain that the sample is random.

Answer: We added a text explaining the sampling process and that it was random (line 182-187).

8. The Authors indicate that the rest of the research results will be published in the following articles. It is important to select appropriate statistical methods in such a case. So far, all of the variables studied are categorical, so they should be placed on the nominal scale in the Stevens’ classification system (Stevens, S.S. On the theory of scales of measurement. Science 1946, 103, 677–680; Stevens, S.S. Psychophysics: Introduction to Its Perceptual, Neural, and Social Prospects; Transaction Publishers: New Brunswick, NJ, USA, 2008; ISBN 978-0-88738-643-5). In this case, the best research methods are the Pearson’s Chi-square test of independence and correspondence analysis.

Answer: Thanks for the suggestion, we will use the appropriate tests for the further analyses.

The article is quite neatly written, contains many interesting details and precise maps of the studied cities with the indication of recreational areas, therefore it brings some added value, although not too great, to the literature on the subject. The only chance to save the text is to treat it as preliminary, pilot research, which should be clearly stated. After clarifying the doubts detailed in the review, there remains the question of the representativeness of the three samples. Using the sample size calculator (https://www.surveysystem.com/sscalc.htm#one) allows you to determine the margin of error (the confidence interval) for each sample with the numbers given in the article. For samples consisting of 235, 255 and 204 units, these margins, for 1 - α = 0.95, are respectively 6.39%, 6.14% and 6.86% (at percentage = 50%). This exceeds the standard values adopted in this type of research (5%), so I leave the decision to publish such an article to the Editor.

Answer: Pilot character is now mentioned in the changed title (line 5).

Reviewer 4 Report

The topic covered in the article is very interesting. I think that the results obtained through the surveys illuminate very well what the citizens want. Also, I think these results can be very useful to public bodies in planning maintenance work on these places and also in planning future new green areas.

I have only minor suggestions for you and I have listed them directly on the attached pdf.

Comments for author File: Comments.pdf

Author Response

The topic covered in the article is very interesting. I think that the results obtained through the surveys illuminate very well what the citizens want. Also, I think these results can be very useful to public bodies in planning maintenance work on these places and also in planning future new green areas.

Answer: Thanks for the appreciation of our work.

I have only minor suggestions for you and I have listed them directly on the attached pdf.

See the file land-1130711-review.pdf

Answer: The minor suggestions were considered and taken completely. We could not change the citation manner since the given one is mandatory for the journal.

Reviewer 5 Report

The manuscript presented for review concerns the topic the value of urban nature in providing ecosystem services for improved health and well-being - an empirical comparative study of three cities in Germany and Czechia. This is one of the "hot topics" in recent years. Recently, more and more attention has been paid to urban nature, ecosystem services, including ecosystem services for improved health and well-being of urban residents. The manuscript is pleasant to read. The tables and graphs included in the manuscript are legible and increase the value of the manuscript.

However, I have a few comments on the manuscript:

1. The title suggests that 3 cities were analyzed in each country - ie 6 cities in total.
2.  Why were two cities not selected in Germany? The characteristics of all city studies should be more emphasized - e.g. in the form of a table. What do these cities have in common (apart from being twin cities) and what divides. Why were these cities chosen and not others?
3.  I believe that photographs from the analyzed cities would greatly contribute to the quality of the manuscript.
4.  Referring to the example of Poland - there is an increase in the number of inhabitants in the suburban area, not necessarily in the city itself. And the suburban area looks more and more like a city than a village. I believe that the authors should also mention this phenomenon, is it also observed in the Czech Republic and Germany?
5. . Did the Authors take into account other statistical tests as well?
6.  The discussion section should be carefully corrected by the Authors - up to line 472 The authors present the description of the results (which should have been included in the previous chapter), and the discussion starts from line 473. Then again the description of the results is mixed with the discussion. I suggest that the Authors consider combining the chapters into results and discussion.
7.  Conclusions in my opinion are too extensive, so it is turdno to make sense. Perhaps the authors would be tempted to have a chapter titled Recommendations? Additionally, there is no more underlined general message from the research carried out. Is it possible to relate the dependencies observed in the 3 cities also to other cities (including cities in other European or world countries)? 

Author Response

Reviewer 5

The manuscript presented for review concerns the topic the value of urban nature in providing ecosystem services for improved health and well-being - an empirical comparative study of three cities in Germany and Czechia. This is one of the "hot topics" in recent years. Recently, more and more attention has been paid to urban nature, ecosystem services, including ecosystem services for improved health and well-being of urban residents. The manuscript is pleasant to read. The tables and graphs included in the manuscript are legible and increase the value of the manuscript.

Answer: Thanks for the appreciation of our work.

However, I have a few comments on the manuscript:

1. The title suggests that 3 cities were analyzed in each country - ie 6 cities in total.

Answer: The title was corrected, the “three” was deleted (line 5).

2. Why were two cities not selected in Germany? The characteristics of all city studies should be more emphasized - e.g. in the form of a table. What do these cities have in common (apart from being twin cities) and what divides. Why were these cities chosen and not others?

Answer: The selection of cities was not done in the study but given by the supporting EU project. We were not allowed to do surveys outside of the three cities using these funds. The description of the green area in the cities were extended, we added each one photo to show some typical characteristics of the cities’ greenery (Figs. 2, 4, 6).

3.  I believe that photographs from the analyzed cities would greatly contribute to the quality of the manuscript.

Answer: This is a good idea, we added each one photo to show some typical characteristics of the cities’ greenery (Figs. 2, 4, 6).

4. Referring to the example of Poland - there is an increase in the number of inhabitants in the suburban area, not necessarily in the city itself. And the suburban area looks more and more like a city than a village. I believe that the authors should also mention this phenomenon, is it also observed in the Czech Republic and Germany?

Answer: This phenomenon was typical in Germany during the 90ies of last century, but the suburbia doesn’t look like a city yet. In Germany, it is actually forbidden to build houses outside the city borderline (in the so-called outside area = “Außenbereich”). Exceptions of this rule are very expensive for the municipalities and rarely happen, therefore.
In the Czech Republic, the situation is rather different from Germany. The suburbanization is not strictly regulated on central/national level. On the local level, municipalities could regulate the increasing of suburban areas by their Municipal Urban Plans. But in practice, increasing of suburban areas is still serious problem of the Czech Republic. Of course, there are some limitations e.g. in national parks or other protected landscape areas where are building closures.

5. Did the Authors take into account other statistical tests as well?

Answer: We did some further statistics were meaningful (section 2, 3) using Chi² test for crosstabs and Spearman-Rho correlation. Additionally, we run several logit models to determine possible relationships between reported feelings and socio-demographic characteristics. The results have been added to the paper (line 322-332, 358-378, 394-422, 446-458, 601-620).

6. The discussion section should be carefully corrected by the Authors - up to line 472 The authors present the description of the results (which should have been included in the previous chapter), and the discussion starts from line 473. Then again the description of the results is mixed with the discussion. I suggest that the Authors consider combining the chapters into results and discussion.

Answer: We would like to maintain the overall structure of the paper since it matches to the rules of the journal. The results mentioned in the discussion section are just repeated from result section in order to discuss them. They should be visible already in results section.

7. Conclusions in my opinion are too extensive, so it is turdno to make sense. Perhaps the authors would be tempted to have a chapter titled Recommendations? Additionally, there is no more underlined general message from the research carried out. Is it possible to relate the dependencies observed in the 3 cities also to other cities (including cities in other European or world countries)? 

Answer: We would like to maintain the overall structure of the paper since it matches to the rules of the journal. But we re-arranged the conclusions to a first part containing knowledge gains and a second part (starting with the clause “From our studies we derive the following recommendations:”, line 795) with conclusions for practice.

We think that most dependencies are similar regarding to other (at least central European) cities, this is shown indirectly by the comparison between German and Czech side. What is stable between that three can be transferred to other cities, too. But since we do not have statistically representative results for more than these three cities, we cannot not claim such generalization in the publication. This must be done by the reader.

Round 2

Reviewer 1 Report

The quality of the manuscript has been improved and significant changes have been made.

However, there are still some questions:

• The authors added information on socio-economic characteristics and other individual information in Table1. However, when reporting data on characteristics, not only the percentage but also the number of participants should be reported and the chi-square test with p-value should be calculated to assess the differences in subject characteristics.

• The relevance of this study was not highlighted in the lines indicated.

Author Response

However, there are still some questions:

• The authors added information on socio-economic characteristics and other individual information in Table1. However, when reporting data on characteristics, not only the percentage but also the number of participants should be reported and the chi-square test with p-value should be calculated to assess the differences in subject characteristics.

Answer: The number of participants are reported in the first two data rows. The percentage relate to them, i.e. each individual number can be calculated easily. We think, to write all numbers in a double way as total and percentage is unnecesary and would make the table confusing. We added the chi-square test in a last column.

• The relevance of this study was not highlighted in the lines indicated.

Answer: Sorry for the confusion with line numbers! The relevance of this study is that we combined the survey on self-sensed health effects, feelings and ecosystem services in a comparison between cities of two European countries, see lines 129-131 in the version with red highlighted changes.

Reviewer 3 Report

The new version of the article is much better than the previous version, which brings the authors closer to publishing the article. However, there are still many errors in the text, and most of them – as before – are methodological in nature. In response to my first review, incorrect line numbers are given, so when assessing the revised version, I was forced to focus on the changes marked in red.

1. The question of the selection of the sample remains unresolved. The article claims that the sample is random. In the new version, it was found that the statistical population includes people who currently use green areas. It is not known whether they are residents of the cities studied, tourists or some other people. This is not a random selection. We can only say that the probability of getting into the sample of a tourist is less than that of a city resident. We deal with random selection when it is possible to calculate the probability of being included in the sample for all units of the statistical population. The Wald-Wolfowitz runs test is used to check the randomness of the sample. There is nothing like this in the article. It should therefore be pointed out that we are dealing with nonprobability sampling. There are situations in social sciences where the sample is selected in a targeted manner by the researcher. In this case, you should try to create a sample that is close to the representative sample. Therefore, it is necessary to include in the text the margins of error for individual samples calculated by me (at the end of the previous review) and explain why the standard value (5%) was exceeded in this type of research. The differences are not that big and are easy to explain. The arguments are presented by the authors in response to point 6 of the previous review, and all these issues should be included in the article. In this way, the reader will find out why we are dealing with a pilot study.
2. The text mentions (line 288) that the Chi-square test was used in the research. What hypotheses were tested? Was it a test of independence? What were the results?
3. The research used Spearman’s Rho which measures the strength and direction of the relationship between two variables. Some of the results are discussed in the text. From the reader’s point of view, however, it is desirable to present a full table of correlation between the studied variables for each city.
4. The conclusions from the research are too general. In response to issue 1 in a previous review, the authors mention that they seek to use their results for more than just gathering scientific knowledge. They believe that their discipline is an applied science, which implies – I guess – some specific practical applications. This is not reflected in the recommendations resulting from the conducted studies (lines 723–742). These recommendations are so general that they can be considered true even without these studies. Examples: 1) places considered dangerous by residents should be protected by the police, 2) the most valuable areas in cities should be designated for green areas, and not used for residential, commercial or transport purposes (by the way: there is no general rule for this, there are only compromises made by local authorities), 3) urban green areas should be close to places where people live, etc. Some issues are not even covered in the questionnaire. So where does this knowledge come from? Why were the maps of the cities studied with the exact location of green areas posted? You have to be consistent in science. Since three cities were studied in detail and the exact locations of green areas were shown on maps, now detailed recommendations for each of these cities have to be provided. Although a few specific examples for each city. Where should additional green areas be placed in each of the examined cities? How can the existing green spaces in Dresden, Děčín and Liberec be connected? Which places related to urban greenery are dangerous for residents? (by the way: this was not asked in the questionnaire). Some recommendations are contradictory. How to reconcile a greater degree of naturalness of green areas, with habitats for plants and animals, with bars and cafes, even of good quality? Natural habitats for selected species of plants and animals are possible only in nature reserves. Besides, these issues are also not included in the questionnaire, so where does this knowledge come from?

Providing specific recommendations for individual cities can help the authorities of these cities in planning the distribution of green spaces, ponds, artificial lakes, etc. Only then can it be said that the research presented in the article is related to applied science.

Author Response

The new version of the article is much better than the previous version, which brings the authors closer to publishing the article. However, there are still many errors in the text, and most of them – as before – are methodological in nature. In response to my first review, incorrect line numbers are given, so when assessing the revised version, I was forced to focus on the changes marked in red.

Answer: We fixed the remaining errors. Sorry for the confusion with line numbers. We had to provide two versions of the text: one with highlighted changes and one without where line numbers differ. The stated numbers relate only to one of the both versions.

1. The question of the selection of the sample remains unresolved. The article claims that the sample is random. In the new version, it was found that the statistical population includes people who currently use green areas. It is not known whether they are residents of the cities studied, tourists or some other people. This is not a random selection. We can only say that the probability of getting into the sample of a tourist is less than that of a city resident. We deal with random selection when it is possible to calculate the probability of being included in the sample for all units of the statistical population. The Wald-Wolfowitz runs test is used to check the randomness of the sample. There is nothing like this in the article. It should therefore be pointed out that we are dealing with nonprobability sampling. There are situations in social sciences where the sample is selected in a targeted manner by the researcher. In this case, you should try to create a sample that is close to the representative sample. Therefore, it is necessary to include in the text the margins of error for individual samples calculated by me (at the end of the previous review) and explain why the standard value (5%) was exceeded in this type of research. The differences are not that big and are easy to explain. The arguments are presented by the authors in response to point 6 of the previous review, and all these issues should be included in the article. In this way, the reader will find out why we are dealing with a pilot study.

Answer: We added a sentence explaning more clearly what the sample should be representative for (lines 156 – 157 in the version with highlighted changes). We deleted the confusing sentence that the sample was random (line 160 in the version with highlighted changes), since randomness has not been tested. We did not ask the people in the green space if they are citizens, students, guests or tourists. Regarding to the purpose of our study we consider this differentiation not important. The improvement of ecosystem services and in particular health effects may apply to citizens similarly to tourists and the conclusions for the maintenance of green spaces should not differ if that would encourage either citizens or people from other places – on the contrary: if an improvement of a urban green spaces‘ quality is not only valuable for citizens but also for tourists and enhance the attractivity of a city should not be a fault. We included the margins of error for individual samples calculated by the reviewer and the arguments are presented by the authors in response to point 6 of the previous review explaining why the standard value (5%) was exceeded in this type of research (line 228-236 in the version with highlighted changes).

2. The text mentions (line 288) that the Chi-square test was used in the research. What hypotheses were tested? Was it a test of independence? What were the results?

Answer: We tested in each case the hypothese of independence of the two factors that have been compared. If the test was positive, we stated that the connection. We added the Chi-square test in Tab. 1 and other statistical values as new appendix (Appendix C) in order to keep the paper compact.

3. The research used Spearman’s Rho which measures the strength and direction of the relationship between two variables. Some of the results are discussed in the text. From the reader’s point of view, however, it is desirable to present a full table of correlation between the studied variables for each city.

Answer: We added statistical values as new appendix (Appendix C) in order to keep the paper compact.

4. The conclusions from the research are too general. In response to issue 1 in a previous review, the authors mention that they seek to use their results for more than just gathering scientific knowledge. They believe that their discipline is an applied science, which implies – I guess – some specific practical applications. This is not reflected in the recommendations resulting from the conducted studies (lines 723–742). These recommendations are so general that they can be considered true even without these studies. Examples: 1) places considered dangerous by residents should be protected by the police, 2) the most valuable areas in cities should be designated for green areas, and not used for residential, commercial or transport purposes (by the way: there is no general rule for this, there are only compromises made by local authorities), 3) urban green areas should be close to places where people live, etc. Some issues are not even covered in the questionnaire. So where does this knowledge come from? Why were the maps of the cities studied with the exact location of green areas posted? You have to be consistent in science. Since three cities were studied in detail and the exact locations of green areas were shown on maps, now detailed recommendations for each of these cities have to be provided. Although a few specific examples for each city. Where should additional green areas be placed in each of the examined cities? How can the existing green spaces in Dresden, Děčín and Liberec be connected? Which places related to urban greenery are dangerous for residents? (by the way: this was not asked in the questionnaire). Some recommendations are contradictory. How to reconcile a greater degree of naturalness of green areas, with habitats for plants and animals, with bars and cafes, even of good quality? Natural habitats for selected species of plants and animals are possible only in nature reserves. Besides, these issues are also not included in the questionnaire, so where does this knowledge come from?

Providing specific recommendations for individual cities can help the authorities of these cities in planning the distribution of green spaces, ponds, artificial lakes, etc. Only then can it be said that the research presented in the article is related to applied science.

Answer: We think, that specific recommendation for the three cities would not be interesting for international readers of the paper (assuming that the readers are rather scientists from all over the world than decision makers from Decin, Liberec and Dresden). Therefore, we collected specific recommendations for individual cities in other documents that are particularly written in the national languages for the authorities of these cities, who hardly read english-speaking scientific journals. These documents are referred as supporting material, mentioned at the end of this paper: „available online at https://bidelin.ioer.eu/ (in German resp. Czech language): The two cited project booklets (for Dresden and Liberec), Study of ES in Děčín, Research report on Assessment of selected ES in Liberec and Děčín, Recommendations for an application of the ES concept in practice, Description of the project, Project Flyer, Press releases, Documentation of several conferences and events, Science pathways, Video on the benefits of urban nature“ (line 759 – 764 in the version with highlighted changes). We added a according reference to the conclusions (line 737 – 739 in the version with highlighted changes).