Addressing Climate Change Impacts on Agriculture: Adaptation Measures For Six Crops in Cyprus

Round 1

Reviewer 1 Report

The paper represents a review of the effectiveness of agricultural measures against climate change harmful effects in a regional scale based on the published data and experts questioning, which is an important topic.

However, the Methods part left several important questions. (1) The questionnaire, which is the base of the whole study, is not described at all. Was the list of possible measures originally proposed in the questionnaire (and respondents were asked to evaluate the effectiveness of each measure, i.e., fill the Table S1) or it was created as a result of questioning (i.e. composed from the measures proposed by respondents). In the 1^st case all respondents should give some evaluation of all measures, in the 2^nd case different measures would be evaluated by different number of experts. In lines 193-199 it was mentioned that some measures were proposed by experts. Does it mean that all others were included by authors themselves? (2) How the proposed measures were graded from 1 to 100 – it was proposed to do in the questionnaire for the respondents (then it will be a set of scores for each measure and criterion) or it was somehow done by authors after analyzing the answers (then it will be only one score for each measure and criterion). (3) How the mean scores were calculated – using whole number of experts or only the number of experts who evaluated this measure. E.g. if only one expert mentioned a certain measure as very good (score 100) and all others did not evaluate it at all – which will be the mean score? (4) It is suggested that the study is based on the extensive literature review. How the data from external literature review was used? For the original selection of possible measures or also for the evaluation of their effectiveness? These external sources evidently should contain more direct numeric data – results of field experiments aimed to increase productivity, economic effectiveness of certain measures, data on crops growing in more hot and dry conditions, e.g., in Israel (which can be used as proxy of future conditions in Cyprus). But no such direct numeric data were applied. I think such data should be mentioned at least in the Discussion.

The higher importance of general measures relatively to the crop-specific ones could be an artifact. I can suppose that most of the experts are not specialists in all six crops, that is why they avoided to evaluate measures for the crops that they are not familiar, whereas general measures, especially such ones as effective irrigation systems and improvement of agricultural advisory, can be evident not only for specialists, but generally for educated people. Maybe it is my mistake, but this issue should be somehow addressed. 

Why such measures as increasing use of recycled wastewater for irrigation or building of desalination facilities (key measures against water shortage in such country as Israel) are not mentioned at all?

Some minor notes are in the attached file 

Comments for author File: Comments.pdf

Author Response

General Note

We re-calculated score mean of each measure using the reversed score for technical difficulty and economic viability criteria (L172-L173) and made the corrections on Table S2 kai S3 (supplementary material). After the corrections, the recommended measures were 12 (instead of 17) and they are given in Table 2. Results and Discussion sections were corrected properly.

Main points

1. METHODS: The questionnaire, which is the base of the whole study, is not described at all. Was the list of possible measures originally proposed in the questionnaire (and respondents were asked to evaluate the effectiveness of each measure, i.e., fill the Table S1) or it was created as a result of questioning (i.e. composed from the measures proposed by respondents). In the 1st case all respondents should give some evaluation of all measures, in the 2nd case different measures would be evaluated by different number of experts.

RESPONSE:  A self-administered semi-structured questionnaire was designed by the research team and used as the basic research tool (L150-151). The 35 identified measures were included in the questionnaire and the respondents evaluated each measure for different characteristics. Respondents were also welcomed to propose their measures at the end of questionnaire, based on an open-ended question. Thus, all the 56 respondents had the opportunity to evaluate the same 35 measures based on the specific 7 criteria. The questionnaire is attached as supplementary material (Table S1). In addition, the number of respondents for each criterion per measure is given in parenthesis in Tables S2 and S3 (supplementary material). Note that the scores for four proposed adaptation measures have been removed from the revised manuscript, because less than 4 respondents had evaluated them.

2. METHODS: In lines 193-199 it was mentioned that some measures were proposed by experts. Does it mean that all others were included by authors themselves?

RESPONSE: Yes. In fact, there were 35 adaptation measures proposed by the research team after the extensive bibliographical research, and 4 additional measures were proposed by the respondents who are experts in the field of climate change.  

3. METHODS: How the proposed measures were graded from 1 to 100 – it was proposed to do in the questionnaire for the respondents (then it will be a set of scores for each measure and criterion) or it was somehow done by authors after analyzing the answers (then it will be only one score for each measure and criterion).

RESPONSE: The evaluation was done in the questionnaire by the respondents. The score of the proposed measures resulted after analyzing the respondents’ replies. It is the average score for each measure and criterion. However, given that less than 4 respondents evaluated them, scores were removed for the revised manuscript.

4. METHODS: How the mean scores were calculated – using whole number of experts or only the number of experts who evaluated this measure. E.g. if only one expert mentioned a certain measure as very good (score 100) and all others did not evaluate it at all – which will be the mean score?

RESPONSE: We used the number of experts that evaluated the measure. Number of people who answered per criterion for each measure are given in parenthesis of Table S2 and S3 in supplementary material. Most experts evaluated all the criteria for the 35 identified measures.

5. METHODS: It is suggested that the study is based on the extensive literature review. How the data from external literature review was used? For the original selection of possible measures or also for the evaluation of their effectiveness?

RESPONSE. The data from external literature review were used only for the original selection of possible measures.

6. METHODS: These external sources evidently should contain more direct numeric data – results of field experiments aimed to increase productivity, economic effectiveness of certain measures, data on crops growing in more hot and dry conditions, e.g., in Israel (which can be used as proxy of future conditions in Cyprus). But no such direct numeric data were applied. I think such data should be mentioned at least in the Discussion.

RESPONSE: The most important references of literature review were added in the manuscript (references 10-47). Indeed, some studies had numeric data and they are given in reference 48 (ADAPT2CLIMA. Review and assessment of national and European legislation, guidelines, plans and best available techniques relative to agriculture; 2016.  Deliverable C2.1. http://adapt2clima.eu/uploads/2017/Del_C2_1.pdf), in which short descriptions of the identified measures are given (pp.32-44).

 For example: 1) “depending on atmospheric conditions, evaporative cooling can reduce canopy temperatures by 8 to 22°C and temperatures of exposed berries by up to 11°C (ref. 33 in our revised manuscript)” (pp 38), 2) “Shading grapevines with neutral density shade cloth (70%) prevents radiative heating by reducing direct radiation and lowers canopy temperatures by approximately 2°C over the growing season, with mid-afternoon canopy temperatures commonly being 4 to 6°C lower than air temperature (ref. 37)” (pp 38), 3) “There is a potential for using kaolin-based particle film, which found to decrease leaf temperature of Cabernet Sauvignon and Merlot varieties by 2 or 3°C (ref. 39) (pp. 39).

However, most data are related to crop-specific not recommended measures (after expert’s evaluation) and, in the Discussion section, we discussed only the recommended measures. For addressing this issue, we added reference 48 in introduction (L125). In addition, the recommended measures are mainly general measures and all of them highly context specific in both space and time. In this sense, presenting numeric data of studies with different experimental conditions (country, plant varieties, climatic conditions, soil conditions etc.) could be misleading (particularly for yield numbers (see also reference 24). However, understanding the need to have representative numbers, in the revised manuscript, we stress the need for conducting sound field experiments in Cyprus for obtaining such appropriate data (L409-L419).

7. RESULTS: The higher importance of general measures relatively to the crop-specific ones could be an artifact. I can suppose that most of the experts are not specialists in all six crops, that is why they avoided to evaluate measures for the crops that they are not familiar, whereas general measures, especially such ones as effective irrigation systems and improvement of agricultural advisory, can be evident not only for specialists, but generally for educated people. Maybe it is my mistake, but this issue should be somehow addressed.

RESPONSE: Indeed, general measures such as “use of efficient irrigation systems and schedules” are already implementing in Cyprus through Rural Development Programme subsidies (L276). On the one hand, this makes experts (and general educated people) familiar with this measure. On the other hand, in Cyprus, farms are small and fragmented, cultivating more than one crop simultaneously and usually a combination of tree and annual crops. Therefore, the more general the measures, the more likely and potentially effective they will be in agricultural holdings with many types of crops. However, for avoiding any misunderstanding, we deleted comparison between the recommended general measures and crop-specific measures in the revised manuscript.

8. RESULTS: Why such measures as increasing use of recycled wastewater for irrigation or building of desalination facilities (key measures against water shortage in such country as Israel) are not mentioned at all?

RESPONSE: In the National Adaptation Strategy of Cyprus, which was one of the most important deliverables developed in the framework of the Adapt2clima project, there is a special chapter on water management, including recycling and water management at River Basin level.

Minor suggestions (in the text)

9. Line 11: “affected by climate change impacts” instead of “affected to climate change impacts”

RESPONSE: We corrected the phrase.

10. Line 15: “measures against drought stress” instead of “measures for drought stress”

RESPONSE: We corrected the phrase.

11. Line 76: “winter?” for potatoes’ yield

RESPONSE: No. The average of all plantings, i.e. winter, spring, and autumn.

12. Line 77: “significantly” instead of “significant”

RESPONSE: We corrected the phrase.

13. Line 94: Unclear. The timing of sowing is not natural phenological issue, but farmer's decision. Why it will delay?

RESPONSE: Indeed, the timing of sowing is farmer’s decision. Thanks for mentioning it. Given the projected advanced in cereal phenology, a yield loss is likely (due to a potential limitation in availability of irrigation water during the critical stages of the plant) if, and only if, farmer has to delay sowing.  In practice, cereal sowing is delayed in the last years in Cyprus, because the first rainfall starts later than in previous years. However, we decided to delete the phrase and rewrite the sentence (L96-97) for avoiding any misunderstanding.

14. Line 126: “ two types?” for the types of climate change impacts

RESPONSE: It refers to the six categories of climate change impacts (L120-L122).  The phrase has been rewritten (L129).

15. Line 136: G is mentioned twice- for wheat and grapes

RESPONSE: We corrected the letter.

16. Line 237: This contradicts Table 3 (for efficiency of cultural practices P=0.651)

RESPONSE: The submitted manuscript refers: L238-239: “However, cultural practices varied significantly in their usefulness, technical difficulty, economic viability and social acceptance (Table 3)”. To our opinion, there is no contradiction. Efficiency does not differ among cultural practices (P=0.651). In contrast, cultural practices varied significantly in their usefulness (P=0.023), technical difficulty (P<0.001), economic viability (P=<0.01) and social acceptance (P=0.010). However, after revised the manuscript, Table 3 was deleted.

17. Line 292: Unclear sentence. Why increase in water demand for crop ET impact on precipitation patterns? Water DEMAND is affected rather by temperature then by precipitation.

RESPONSE: The sentence had been rewritten L266-L268: “Climate change impacts on precipitation patterns and their distribution through the year increase the demand for water, due to increased crop evapotranspiration, and increase water shortages, particularly in the spring and summer months”.

18. Line 298: “aim to improving” instead of “aim at improving”

RESPONSE: We corrected the phrase – Line 272.

Reviewer 2 Report

The manuscript entitled "Addressing climate change impacts on agriculture: adaptation measures for six crops in Cyprus" analyses expert opinion on how to adapt agriculture to climate change from a general point of view and for specific types of crops that are hot spots for this problem. The manuscript is very well written, except for a small problem of lack of explanation in materials and methods that will be easily overcome.

In material and methods it says that each criterion was evaluated from 1 to 100, but the results of the statistical analysis (line 215 to 220) says that N=925, so if I understand correctly, it will be that it was evaluated from 1 to 1000. Please explain better or correct.

In figure 1 and 2, the numbers seem to be outlayers but it should be indicated in the legend

The basic problem of this manuscript is, in my view, the contribution to the journal Atmosfere. Reading aims and score of it, I do not find that anything related to the atmosphere is analyzed.The closest topic is the analysis of anthropogenic effects of changes in use, which this manuscript does not analyze either. I would suggest that it be sent to the journal Sustainability or one more focused on Agriculture.

Author Response

General Note

We re-calculated score mean of each measure using the reversed score for technical difficulty and economic viability criteria (L172-L173) and made the corrections on Table S2 kai S3 (supplementary material). After the corrections, the recommended measures were 12 (instead of 17) and they are given in Table 2. Results and Discussion sections were corrected properly.

Main Points

1. In material and methods it says that each criterion was evaluated from 1 to 100, but the results of the statistical analysis (line 215 to 220) says that N=925, so if I understand correctly, it will be that it was evaluated from 1 to 1000. Please explain better or correct.

RESPONSE: The evaluation of each criterion was from 1 to 100 (L154-162). In the statistical analyses, N refer to records. For example N=925 means that we had 925 answers for the efficiency of the recommended measures. However, we decided to remove statistical analysis of the revised manuscript because it seems to have no added value for our results.

2. In figure 1 and 2, the numbers seem to be outlayers but it should be indicated in the legend.

RESPONSE: Numbers have been removed in Fig 2 (which is Fig.1 in revised manuscript) and Y-axis legend is added. Numbers were refer to SPPS record line of the outlier (not the outlier value) and should have been removed. Thanks for mentioning it.

3. The basic problem of this manuscript is, in my view, the contribution to the journal Atmosphere. Reading aims and score of it, I do not find that anything related to the atmosphere is analyzed. The closest topic is the analysis of anthropogenic effects of changes in use, which this manuscript does not analyze either. I would suggest that it be sent to the journal Sustainability or one more focused on Agriculture.

RESPONSE:  The paper has been prepared for a Special Issue, namely “Adaptation of Cyprus Agriculture to Climate Change”, of the section: Biometeorology of the ATMOSPHERE journal.

Reviewer 3 Report

In the article, the authors evaluate and discuss measures that can be implemented in the agriculture of Cyprus to mitigate the impact of climate change on six major crops.

Review:

L76: Why did you mention frost? How does it relate to the manuscript topic i.e. climate change? Is it an issue in Cyprus?

L91: Please provide references for all models.

L94: Please explain why cereal sowing in Cyprus would be delayed due to climate change?

L115: Please provide references for the main sources of measures.

Have you tried to estimate the measures based on the models listed in the Introduction section, e.g. the efficiency of the measure based on the yield obtained? Some of these measures can be implemented in these growth models. Maybe it would be good to comment on that.

L145: please clarify 0 and 100 for technical difficulty (later it is explained that 0 is easy and 100 – difficult).

L152: Please justify the use of both statistical tests and describe them more clearly.

As far as I can see, the steps are as follows:

1. each expert evaluated each of the 35 measures presented in Table 1 for each criterion, giving it a number on a scale of 0 to 100
2. you chose best measures in each category of measures based on the total score
3. for some measures, you carry out the Kruskall-Wallis test for each measure category to compare the measures for each criterion
4. for the measures selected in point 2, you carry out the Mann-Whitney test to compare general measures with crop-specific ones in each criterion?

In my opinion, the whole procedure requires additional clarification. Why did you compare general measures with crop-specific ones? How important is that?

L215: I think you could delete U,N and H values, the p-value gives enough information. The rest might may not be necessary for the reader.

Fig. 1 and 2. Please add the axes and their units. What do the numbers in the graph mean?

Is statistical analysis necessary? It seems to me that the discussion is mainly based on Table 2, and the test results are not discussed.

Table S1 there is no SD only means. As far as I can see, the “Total score” was calculated as the arithmetic mean, why? Are all criteria equally important?

The idea of ​​this work is good and relates to a very important topic. The study, however, is based on expert judgement. This can be subjective. It would be good to support the study with some references, in which these measures were assessed using certain numbers. The authors wrote that they conducted a comprehensive literature review, which I did not see in the text. In addition, a statistical analysis was performed that is not properly described and discussed. Please consider whether the work would not be more clear without this analysis.

Specific comments:

Please correct and unite the units for tonnes per hectare.

Check all numbers in the text, they do not match the numbers in Table S1 e.g.

L170: It should be 78%

L177: It should be 85% and so on…

Author Response

Main points

1. L76: Why did you mention frost? How does it relate to the manuscript topic i.e. climate change? Is it an issue in Cyprus?

RESPONSE: Frost is not a common extreme event in Cyprus (such as heat stress), however whenever it happens can cause potatoes’ yield loss. In this sense, we mentioned the frost as a fact leading to reduced yields in potatoes and only in the introduction of the manuscript, as an example of extreme weather events.

2. L91: Please provide references for all models.

RESPONSE: Corrected. References 6-8 (L95).

3. L94: Please explain why cereal sowing in Cyprus would be delayed due to climate change?

RESPONSE: Given the projected advanced in cereal phenology, a yield loss is likely (due to a potential limitation in availability of irrigation water during the critical stages of the plant) if, and only if, farmer has to delay sowing.  In practice, cereal sowing is delayed in the last years in Cyprus, because the first rainfall starts later than in previous years. However, we decided to delete the phrase and rewrite the sentence (L96-97) for avoiding any misunderstanding.

4. L115: Please provide references for the main sources of measures.

RESPONSE: Corrected. References 10-47 are related to literature review (L119). In addition, a short description of the identified measures can be found in reference 48 (ADAPT2CLIMA. Review and assessment of national and European legislation, guidelines, plans and best available techniques relative to agriculture; 2016.  Deliverable C2.1. http://adapt2clima.eu/uploads/2017/Del_C2_1.pdf) (pp.23-44) (incl. bibliography). Note that questionnaire survey was conducted during the September of 2016 (L118), so the literature review used is updated until 2016.

5. Have you tried to estimate the measures based on the models listed in the Introduction section, e.g. the efficiency of the measure based on the yield obtained? Some of these measures can be implemented in these growth models. Maybe it would be good to comment on that.

RESPONSE: We included only irrigation and shifting of sowing dates (sowing crop) and variety (for olive trees and grapes) in growth models.

6. L145: please clarify 0 and 100 for technical difficulty (later it is explained that 0 is easy and 100 – difficult).

RESPONSE: Corrected (Line 158).

7. L152: Please justify the use of both statistical tests and describe them more clearly. As far as I can see, the steps are as follows:
   • each expert evaluated each of the 35 measures presented in Table 1 for each criterion, giving it a number on a scale of 0 to 100
   • you chose best measures in each category of measures based on the total score
   • for some measures, you carry out the Kruskall-Wallis test for each measure category to compare the measures for each criterion
   • for the measures selected in point 2, you carry out the Mann-Whitney test to compare general measures with crop-specific ones in each criterion?

In my opinion, the whole procedure requires additional clarification. Why did you compare general measures with crop-specific ones? How important is that?

RESPONSE: Taking into consideration the above issue, the methodology (2.2. Evaluation of the identified adaptation measures L145-176) has been rewritten in order to make steps clear. In addition, the statistical analyses (based on non-parametric comparisons of mean values for two or more than two measures) have been deleted in the revised manuscript, since we believed that do not have added value (particularly after re-calculating the average score according to the methodology described above L170-173).

In addition, we have deleted the comparison between the recommended general measures and the crop-specific measures. The first thought was to highlight that general measures are potentially more appropriate for Cypriot farmers whose farms are small and fragmented, cultivating more than one crop simultaneously and usually a combination of tree and annual crops. Therefore, the more general the measures, the more likely and potentially effective they will be in agricultural holdings with many types of crops. On the other hand, the recommended measures resulted from the subjective experts’ judgment, so our justification may be false.

8. L215: I think you could delete U, N and H values, the p-value gives enough information. The rest might may not be necessary for the reader.

RESPONSE: Statistical analyses have been deleted.

9. Fi 1 and 2. Please add the axes and their units. What do the numbers in the graph mean?

RESPONSE: Figure 1 (submitted manuscript) has been deleted. Numbers have been removed in Fig 2 (which is Fig.1 in revised manuscript) and Y-axis legend is added. Numbers were refer to SPSS record line of the outlier (not the outlier value) and should have been removed.

10. Is statistical analysis necessary? It seems to me that the discussion is mainly based on Table 2, and the test results are not discussed.

RESPONSE: Thanks for the comment. After taking into consideration this issue, we decided to remove the statistical analyses.

11. Table S1 there is no SD only means.

RESPONSE: Corrected. SD were added in all means numbers. See Table S2 and S3 in the

 Supplementary material.

12. As far as I can see, the “Total score” was calculated as the arithmetic mean, why? Are all criteria equally important?

RESPONSE: Thank you for this observation. Indeed, the score for each measure in Table 2 of the manuscript was calculated as the arithmetic mean or, in other words, as the weighted average (using equal weights) of the score of each criterion. The assignment of equal weights to the criteria was decided by the research team of the project ADAPT2CLIMA in Cyprus after several discussions, as it was considered that all seven criteria are of equal importance. For instance, the research team unanimously agreed that social acceptance of a measure is equally important with economic viability and vice versa. In addition, equal weighting is the method most commonly used when no explicit evidence to prefer a criterion over another is available, as is the case with this study. We acknowledge that the calculation process is not appropriately described in the manuscript; thus we revised the text to make things more clear (L145-174). In addition, we re-calculated score mean of each measure using the reversed score for technical difficulty and economic viability criteria (L172-L173) and make the corrections on Table S2 kai S3 (supplementary material). After the corrections, the recommended measures were 12 (instead of 17) and they are given in Table 2. Results and Discussion sections were corrected properly.

13. The idea of this work is good and relates to a very important topic. The study, however, is based on expert judgement. This can be subjective. It would be good to support the study with some references, in which these measures were assessed using certain numbers. It would be good to support the study with some references, in which these measures were assessed using certain numbers.

RESPONSE: Thank you for this comment. We agree with the reviewer that the overall assessment is based on the subjective judgement of experts and this may be considered a limitation. However, such an approach is a pragmatic and fast way of assessing complex phenomena (Zahm et al., 2008; Markou et al., 2020), such as climate change adaptation, in a cross-sectional study such as the one presented in this manuscript.

Following your suggestions, the most important references of literature review were added in manuscript (references 10-47) (L120). Indeed, some studies had numeric data and they are also given in reference 48 (ADAPT2CLIMA. Review and assessment of national and European legislation, guidelines, plans and best available techniques relative to agriculture; 2016.  Deliverable C2.1. http://adapt2clima.eu/uploads/2017/Del_C2_1.pdf), in which short descriptions of the identified measures are given (pp.32-44).

 For example: 1) “depending on atmospheric conditions, evaporative cooling can reduce canopy temperatures by 8 to 22°C and temperatures of exposed berries by up to 11°C (ref. 33 in our revised ms)” (pp 38), 2) “Shading grapevines with neutral density shade cloth (70%) prevents radiative heating by reducing direct radiation and lowers canopy temperatures by approximately 2°C over the growing season, with mid-afternoon canopy temperatures commonly being 4 to 6°C lower than air temperature (ref. 37 in our revised ms)” (pp 38), 3) “There is a potential for using kaolin-based particle film, which found to decrease leaf temperature of Cabernet Sauvignon and Merlot varieties by 2 or 3°C (ref. 39 in our revised ms) (pp. 39).

However, most numeric data are related to crop-specific not recommended measures (after expert’s evaluation). Given that in the Discussion section, we discussed only the recommended measures, we added reference 48 in Introduction (L125).

14. The authors wrote that they conducted a comprehensive literature review, which I did not see in the text.

RESPONSE: See the answer of Q4.

15. In addition, a statistical analysis was performed that is not properly described and discussed. Please consider whether the work would not be more clear without this analysis.

RESPONSE: Please see the answers of Q7 and Q10.

 SPECIFIC COMMENTS

16. Please correct and unite the units for tonnes per hectare.

RESPONSE: Corrected

17. Check all numbers in the text, they do not match the numbers in Table S1 e.g. L170: It should be 78% L177: It should be 85% and so on…

RESPONSE: All means in the text have been checked and corrected. All means of Tables S2 and S3 have been checked and corrected, and “score” was re-calculated accordingly to the methodology described in the revised manuscript, considering the revised score of technical difficulty and economic viability criteria (L145-L175).

Round 2

Reviewer 1 Report

The authors dramatically improved Methods part, clearly describing the questionnaire and its application. The authors also mentioned in Conclusions the main limitation of the study (based on subjective evaluations of experts), which is also was good to add.

I have one major note. In the answer 8 (about recycled wastewater for irrigation or building of desalination facilities) the authors wrote:

RESPONSE: In the National Adaptation Strategy of Cyprus, which was one of the most important deliverables developed in the framework of the Adapt2clima project, there is a special chapter on water management, including recycling and water management at River Basin level.

However, from my opinion it is not a response, why these measures are not mentioned at all, as other measures, which are included in the study (e.g., advanced irrigation techniques) are also mentioned in other projects (what is stated in answer 7). If the authors wanted to focus only to the local measures, excluding country-scale ones (like desalination), you should somehow mention it in the Methods.

Author Response

Dear reviewer,

In reply to your comment, we have included lines 118-125 in the methodology section, where we explain why the use of recycled or desalinization water was not included as adaptation measure in our study.

Thank you