Assessment of climate change impact on the malaria vector Anopheles hyrcanus, West Nile disease, and incidence of melanoma in the Vojvodina Province (Serbia) using data from a regional climate model

Motivated by the One Health paradigm, we found the expected changes in temperature and UV radiation (UVR) to be a common trigger for enhancing the risk that viruses, vectors, and diseases pose to human and animal health. We compared data from the mosquito field collections and medical studies with regional climate model projections to examine the impact of climate change on the spreading of one malaria vector, the circulation of West Nile virus (WNV), and the incidence of melanoma. We analysed data obtained from ten selected years of standardised mosquito vector sampling with 219 unique location-year combinations, and 10 years of melanoma incidence. Trends in the observed data were compared to the climatic variables obtained by the coupled regional Eta Belgrade University and Princeton Ocean Model for the period 1961–2015 using the A1B scenario, and the expected changes up to 2030 were presented. Spreading and relative abundance of Anopheles hyrcanus was positively correlated with the trend of the mean annual temperature. We anticipated a nearly twofold increase in the number of invaded sites up to 2030. The frequency of WNV detections in Culex pipiens was significantly correlated to overwintering temperature averages and seasonal relative humidity at the sampling sites. Regression model projects a twofold increase in the incidence of WNV positive Cx. pipiens for a rise of 0.5°C in overwintering TOctober–April temperatures. The projected increase of 56% in the number of days with Tmax ≥ 30°C (Hot Days—HD) and UVR doses (up to 1.2%) corresponds to an increasing trend in melanoma incidence. Simulations of the Pannonian countries climate anticipate warmer and drier conditions with possible dominance of temperature and number of HD over other ecological factors. These signal the importance of monitoring the changes to the preparedness of mitigating the risk of vector-borne diseases and melanoma.

Authorship changes: Aleksandra Ignjatović Ćupina (AIC) contributed to the acquisition of the Anopheles hyrcanus data and has been added to the list of authors. All authors are informed and express agreement regarding this change.
Introduction L49-53: Are the authors referring to themselves when they stated, "The authors (……….), have been working together……."? Yes, we tried to make it clear with correction below. L56: Corrected to: "The authors of the manuscript (……….), have been working together……." Results -Authors should specify the exact p-values instead of writing p<0.05 or p>0.05 L260-281: Corrected according to the suggestion. The section was updated to include the exact p-values.
L207: "The Poisson regression model for the dependence of a number of detections per site (frequency-λ)…………………………is highly significant". Authors stated it was highly significant, but from my perspective, p<0.05 is not a specific indication of high significance. Could the give the exact value of p? L282-283: Corrected according to the suggestion. Exact p value is not below 0,01 (it is 0.01393) which is considered as high significance by many authors, so we erased the word "highly".
-Fig2b and 2c are fuzzy Thank you for your comment. The figures were reformatted to higher resolution according to PACE. It will be more interesting if the authors used only vector-borne diseases data in this paper. N.B: Other comments are incorporated in the manuscript Authors appreciate very much the effort invested in the improvement of the manuscript quality. All suggestions are incorporated into the revised version except one concerning the spelling of NUTS. Nomenclature of territorial units for statistics is originally abbreviated NUTS from the French version (Nomenclature des Unités territoriales statistiques).
Reviewer #2: Authors are presenting an interesting paper regarding the effects of climate change in Northern Serbia considering three independent measures: The spread of Anopheles hyrcanus, the presence of West Nile Virus in Culex pipiens, and the incidence of melanoma cases. The paper is interesting, however, discussion should be improved specially on the uncertainty of future predictions since they are using just one climatic model. Further, their results should be stated more carefully since their model rely on assumptions (e.g., manually selected variables) which are also not clearly stated. Discussion is corrected according to the reviewer's suggestion: L345-352: New text added: The temperature trend over the period of observations used in this study and for the future time horizon following A1B scenario obtained with the EBU-POM regional climate model is within the multi-model ensemble (MME) spread of regional climate models with similar configuration used in the ENSEMBLES project (van der Linden and Mitchell, 2009). For the period 2001-2030 the temperature change for the region of interest in the EBU-POM integration is 0.75 C concerning the period 1961-1990 and for the same period ENSEMBLES MME spread range is 0.5-1.5 C (MEP, 2017). Following this finding, other results presented in this paper that relay on temperature change, can be seen as an estimate that will be within uncertainty related to the future temperature projection.
New references included: van der Linden P and Mitchell JFB, editors. ENSEMBLES Climate Change and its Impacts. Summary of research and results from the ENSEMBLES project. Authors are using one of the SRES future climatic scenarios; currently the standard for future climate studies are the RCP scenarios. Authors should describe the nature of the SRES-A1B scenario, which is not mentioned in any part of the study. Further, authors should explicitly discuss uncertainty on their predictions since they are not using other scenarios or other climatic models. Authors addressed this comment in the text corrected. Please check response to the L113-122 and L 345-352 above.
Lines 176-180. There is no discussion or results regarding these sentences. Was the comparison between EBU-POM model and the Republic Hydrometeorological Service of Serbia perfect? What is the implication of this approach on the overall paper? This is a valuable comment since the information measure(s) is(are) a good indicator of the reliability of model outputs and thus on the overall paper. The increasing complexity of climate models is a growing concern in the modelling community. However, because we invested a serious effort to make our models more "realistic", we included more parameters and processes. With increasing model complexity, we are less able to manage and understand the model behaviour. Thus, from a user's perspective, the following question is entirely natural: "How complex model (EBU-POM model in our case) do I need to use to study this problem (assessment of climate change impact on malaria vectors, West Nile disease, and incidence of melanoma in the VPS) with this data set (temperature and/or precipitation)?". In the revised version, we inserted the additional text. L229-249: New text added: We considered the papers by Mihailović et al. [2,24] in which Kolmogorov complexity measures (Kolmogorov complexity (KC), Kolmogorov complexity spectrum KC spectrum) and the highest value of the KC spectrum (KCM)) and sample entropy (SE) [25] were used to quantify the regularity and complexity of air temperature and precipitation time series, obtained by the EBU-POM model, representing both deterministic chaos and stochastic processes. We considered the complexity of the EBU-POM model using the observed and modelled time series of temperature and precipitation. We computed the KC spectrum, KC, KCM and SE values for temperature and precipitation. The calculations were performed for the entire time interval 1961-1990: (1) on a daily basis with a size of N =10958 samples for temperature and (2) on a monthly basis with a size N =360 for the precipitation. The simulated time series of temperature and precipitation were obtained by the EBU-POM model for the given period. The observed time series of temperature and precipitations for two stations: Sombor (SO) (88 m a.s.l.) and Novi Sad (NS) (84 m a.s.l.) in the considered area, were taken from daily meteorological reports of the Republic Hydrometeorological Service of Serbia. For both sites, the modelled complexity is lower than the observed one, but with the reliability which is in the interval values allowed by the information measures (KC, KCM, and SE) (Krzic et al. 2011, Dell' Aquila et al. 2016, Cavicchia et al. 2016). These findings mean that the models with a KC (and KCM) complexity lower than the measured time series complexity cannot always reconstruct some of the structures contained in the observed data. However, it does not mean that outputs from EBU-POM model do not correctly simulate climate elements since both sites values indicate the absence of stochastic influences, providing reliable projections of the climate elements. New references included: Krzic  Line 277-280: There is no evidence in this paper supporting this affirmation since the variables analyzed corresponded to three temperature related variables and just one considering humidity. Moreover, results were never compared statistically; modify accordingly. Corrections made as suggested. The sentence "It seems that temperature in semiurban areas dominates the other environmental factors influencing WNV circulation in nature (e.g. landscape suitability for reservoir host and mosquito vector, host availability, precipitation), as it is the primary factor affecting both mosquito vector abundance and virus replication." now reads as: L400-404: Corrected to: It seems that temperature in semi-urban areas is an essential environmental factor influencing WNV circulation (landscape suitability for reservoir host and mosquito vector, host availability, and precipitation/water availability are somewhat similar in investigated semi-urban areas of VPS), as it affects both mosquito vector abundance and virus replication.  : Add WD and HD to the corresponding legend of the graphic. Is there a Croatian sentence in the legend? Please describe how the melanoma incidence was calculated, is the y axis showing incidence or number of cases? Cumulative incidence is known to over-represent trends (see reference: Vandenbroucke & Pearce, 2012, doi: 10.1093/ije/dys142), try to use incidence rate instead. This is a keystone issue in this field of epidemiology. However, it is still under a broad umbrella of discussion. In particularly mentioned reference (Vandenbroucke & Pearce, 2012) the Authors comprehensively considered the place of incidence rates in dynamic populations as well as the cumulative incidence (risk or portion) from an epistemological point of view and also giving very illustrative (educational) examples. Many authors were arguing with some ideas explicated in this paper, also considering some examples. We agree with V&P ideas, but we did not find the place where they explicitly say that it would always be using the incidence rate instead of cumulative (the question of overestimation). To our understanding, they left the space for a situation when the use of cumulative incidence gives acceptable results. For example, it is partly seen in the paper by Wu et al. (2014). There is another moment why we used cumulative incidence. It is well-known that there is a high correlation between sun exposure (and received cumulated doses of the UV radiation) and melanoma. If that doses (or any climate element) on a daily basis are used from regional climate models, they cannot be directly correlated with daily or monthly measured or calculated biological quantities. The reason for that is the fact that from regional climate models, we can estimate just the trend of the considered physical quantity (in our case -UV doses through their cumulative values). Correspondingly it is correlated with cumulative incidence. Having said that, after the end of the statement in Line 336, we inserted the following text. The legend in Fig. 4(c) and y-axis in Fig 4 (d) are changed as suggested. The M&M -Melanoma incidence and UVR was amended by the following text: L216-221: New text added: In the analysis we have used two indicators: (i) melanoma incidence rate that is a measure of the number of new cases ("incidence") per unit of time ("rate") and (ii) cumulative incidence ("incidence proportion" that measures the number of new cases per person in the population over a defined period of time -often called risk or proportion). Melanoma incidence rate (per 100,000 people) for ten years 1995 -2004 was based on the data obtained from the paper by Jovanović et al. [7]. From these data, we calculated the cumulative incidence. The discussion was also amended by the following text: L437-441: New text added: In a cohort study, Wu et al. (2014) considered the impact of long-term UV radiation flux on skin cancer risk. Comparing with participants in the lowest quintile of cumulative UV radiation flux in adulthood, they found that participants in the highest quintile had multivariable-adjusted risks (cumulative incidence). According to Vandenbroucke and Pearce (2012), some studies where cumulative incidence is used can over-represent the trends. New references included: Wu Figure 6 can be replaced with the statistics of such graphic for readers' interpretation. L442-446: Corrections made as suggested. Figure 6 deleted, the paragraph now reads as: From a statistical point of view, the linear regression model for modelling the cumulative incidence of melanoma versus the difference of the cumulative UVR doses for hot and warm days (Fig 4d) is acceptable. Parameters are statistically highly significant (r = 0.9712 and p = 0.000003) while analysis of residual distribution shows a good agreement with the normal distribution (Shapiro-Wilk test, W = 0.9608, p = 0.7952).
Authors are justifying the paper under the 'One Health' concept, however they are not discussing the idea further. I would like discussing explicitly the benefits of putting together a set of multidisciplinary specialists to the development of the manuscript and how this contribution is part of the one health concept. The discussion was amended by the following text: L322-344: New text added: Despite globalisation trends, researchers have become "closed" in their ever-smaller communication circles which are not limited by state or national borders but by the professional language and way of thinking. Thus, by the end of the 20th century, the scientific community has been faced with problems in communication within its confines. One of the principal reasons why vector-borne diseases (VBD) are so difficult to predict, is the complex interaction of multiple factors (vector, host, pathogen, environment including short-term weather patterns and long-term climate change) in space and time . Only groups from multiple sectors that communicate and work together on specific aspects of VBD systems will be able to answer the most exciting and pressing problems in the field (Moore 2008). Authors of this paper started collaboration in 2003 comparing the climates of the foci of WNV circulation in USA (California Central Valley) and Europe (Bucharest area) with VPS. As compared climates showed quite similar patterns, colleagues from public health and veterinary joined the initial group of meteorologists and medical entomologists. With the idea to better draw upon the resources and insights of the various sectors we designed and implemented research and programmes to achieve better outcomes in the control of zoonoses (diseases that can spread between animals and humans, e.g., WNV disease). This led us to the following achievements: (i) the first detection of WNV in horses in Serbia in 2009 (Lupulović 2011); (ii) the first detection of WNV in mosquitoes in Serbia in 2010 (26); (iii) the first detection of WNV in wild birds in Serbia in 2012 (Petrović 2013 Minor comments: Please use Oxford comma across the manuscript: e.g., Line 30: 'the malaria vector, and the incidence of melanoma'. Corrections made as suggested. Line 28: Authors never discuss problems related with animal health, thus, I suggest avoiding this kind of affirmations (e.g., line 81). The reviewer is right, we did not, but we think it is vital to mention animals because WNV is the important zoonotic diseases. Therefore, we would like to include new paragraphs in Introduction and Discussion. The introduction was amended by the following text: L76-80: New text added: In Europe, the total number of reported autochthonous WNV infections in 2018 (n=2,083) exceeded, by far, the total number from the previous seven years (n=1,832 The WNV transmission cycle involves mosquito vectors and birds, but equines and humans are also susceptible to infection , Blitvich 2008. Although WNV infections have been described in a wide variety of vertebrates, birds are the main natural reservoir. Hundreds of wild and domestic avian species have been described as susceptible to WNV infection, but many of these showed only subclinical infection . In horses, WNV infection is also frequently clinically unapparent, but around 10% of cases develop neurological disorders with up to 50% mortality rates (Blitvich 2008). An increasing number of severe outbreaks in horses have been reported in Europe in the past decade, including a large one that took place in northeast Italy in 2008 involving 251 stables with 794 cases and five deaths   According to the suggestion, the text placed between 172-174 lines is replaced by the following one. L216-221: New text added: In the analysis we have used two indicators: (i) melanoma incidence rate that is a measure of the number of new cases ("incidence") per unit of time ("rate") and (ii) cumulative incidence ("incidence proportion" that measures the number of new cases per person in the population over a defined period of time -often called risk or proportion). Melanoma incidence rate (per 100,000 people) for ten years 1995 -2004 was based on the data obtained from the paper by Jovanović et al. [14]. From these data, we calculated the cumulative incidence.
Line 227: Is the formula correct: warm days -WD? L301-302: Changed to: air temperature Tmax ≥ 25 C (Warm Days -WD) Line 263: Consider changing 'indicate that the findings supporting' by 'support' L370-374: The sentence was quietly confusing; we rewrote it to read like this: Positive trends which are present in our observations might indicate that the findings on the negative influence of UVR and blue-light radiation (this radiation has a wavelength between approximately 380 nm and 500 nm; it has a very short wavelength, and so produces a higher amount of energy) on adult mosquitoes under laboratory conditions [38,39] could not be simply translated to the field.   Motivated by the One Health paradigm, we found the expected changes in temperature and UV radiation 29 (UVR) to be a common trigger for enhancing the risk that viruses, vectors, and diseases pose to human 30 and animal health. We compared data from the mosquito field collections and medical studies with 31 regional climate model projections to examine the impact of climate change on the spreading of one 32 malaria vector, the circulation of West Nile virus (WNV), and the incidence of melanoma. We analysed 33 data obtained from ten selected years of standardised mosquito vector sampling with 219 unique location-34 year combinations, and 10 years of melanoma incidence. Trends in the observed data were compared to 35 the climatic variables obtained by the coupled regional Eta Belgrade University and Princeton Ocean 36 In this paper, the authors collected and analysed observed data over 31 years and related a subset to 58 outputs from a Regional Climate Model (RCM). Vector-borne diseases and melanoma are significant 59 climate-driven threats for which risk sources can be clearly defined [6]. In this study, devoted to revealing the potential impact of climate change on animal and human health, we 89 compared a considerable amount of previously unpublished ecological data obtained from the field and 90 clinical surveys with climate change projections for the VPS, which is representative of the Central 91 European low-altitude areas with a human-dominated landscape (Fig 1). We examined the effects of 92 temperature on the spread and relative abundance of the malaria vector An. hyrcanus and the 93 "microclimate" differentiation between sites with a specific frequency of WNV occurrence in Cx. pipiens 94 . We also evaluated the impact of climate change on melanoma incidence as a synergy of changes in UVR 95 doses and the long-term increase in the number of hot days (HD), with daily maximum temperature ≥ 96 30 o C using the Eta Belgrade University and Princeton Ocean Model (EBU-POM) regional model data.  Table).

201
In the analysis we have used two indicators: (i) melanoma incidence rate that is a measure of the number 202 of new cases ("incidence") per unit of time ("rate") and (ii) cumulative incidence ("incidence proportion" 203 that measures the number of new cases per person in the population over a defined period of timeoften 204 called risk or proportion). Melanoma incidence rate (per 100,000 people) for ten years 1995 -2004 was 205 based on the data obtained from the paper by Jovanović et al. radiation flux in adulthood, they found that participants in the highest quintile had multivariable-adjusted 403 risks (cumulative incidence). According to Vandenbroucke and Pearce [71], some studies where 404 cumulative incidence is used can over-represent the trends. 405 From a statistical point of view, the linear regression model for modelling the cumulative incidence of 406 melanoma versus the difference of the cumulative UVR doses for hot and warm days (Fig 4d) is 407 acceptable. Parameters are statistically highly significant (r = 0.9712 and p = 0.000003) while analysis of 408 residual distribution shows a good agreement with the normal distribution (Shapiro-Wilk test, W = 409 0.9608, p = 0.7952). 410 We hope that our results will indicate the importance of long-term monitoring/surveillance programs for 411 providing crucial data to evidence the ongoing biological alteration triggered by climate change. 412 Nonetheless, it is difficult to say how broadly our data represent the trends elsewhere. We believe that the 413 specificity of the observations offers a unique window into the state of some of the planet's pressing 414 threats to human health. Also, in the case where humans are exposed to UVR, due to the nature of their 415 work (the VPS is an exclusively agricultural area), it is necessary to (i) establish a broader network for 416 UVR measurements and warning centres and (ii) increase the awareness of the melanoma as a result of 417 Motivated by the One Health paradigm, we found the expected changes in temperature and UV radiation 30 (UVR) to be a common trigger for enhancing the risk that viruses, vectors, and diseases pose to human 31 and animal health. We comparecompared data from the mosquito field collections and medical studies 32 with regional climate model projections to examine the impact of climate change on the spreading of one 33 malaria vector, the circulation of West Nile virus (WNV), the spreading of the malaria vector and the 34 incidence of melanoma. We analysed data obtained from ten selected years of standardizedstandardised 35 mosquito vector sampling with 219 unique location-year combinations, and 10years of melanoma 36 incidence. Trends in the observed data were compared to the climatic variables obtained by the coupled 37 In this paper, the authors collected and analysed observed data collected over a period of 31 years and 61 related a subset to outputs from a Regional Climate Model (RCM). Vector-borne diseases and melanoma 62 are significant climate-driven threats for which risk sources can be clearly defined [6]. In this study, devoted to revealing the potential impact of climate change on animal and human health, we 96 comparecompared a considerable amount of previously unpublished ecological data obtained from the 97 field and clinical surveys with climate change projections for the VPS, which is representative of the 98 Central European low-altitude areas with a human-dominated landscape (Fig 1). We examined the effects 99 of temperature on the spread and relative abundance of the malaria vector An. hyrcanus and the 100 "microclimate" differentiation between sites with a specific frequency of WNV occurrence in Cx. pipiens 101 and effects of temperature on the spread and relative abundance of the malaria vector An. hyrcanus.. We 102 5 also evaluated the impact of climate change on melanoma incidence as a synergy of changes in UVR 103 doses and the long-term increase in the number of hot days (HD), with daily maximum temperature >≥ 104 30 o C using the Eta Belgrade University and Princeton Ocean Model (EBU-POM) regional model data. is seen in FigFigs 1a and Fig 1b). This region is the essential food production area in Serbia with a total 126 6 surface area of 21,500 km 2 and a population of about 2 million. This region has a continental climate, 127 with elements of a sub-humid and warm climate (Cfwbx" according to Köppen classification).

Models and formula used 129
The global and regional climate model  Table). In the analysis we have used two indicators: (i) melanoma incidence rate that is a measure of the number 216 of new cases ("incidence") per unit of time ("rate") and (ii) cumulative incidence ("incidence proportion" 217 that measures the number of new cases per person in the population over a defined period of time -often 218 called risk or proportion). Melanoma incidence rate (per 100,000 people) for ten years 1995 -2004 was 219 based on the data obtained from the paper by Jovanović et al. [7]. From these data, we calculated the 220 cumulative incidence. We have used the model simulation to study the expected impact of climate change The work presented in paper Mihailović et al. is interesting. The objective of the authors was to compare data from the mosquito field collections and medical studies with regional 29 climate model projections to examine the impact of climate change on the circulation of West Nile virus (WNV), the spreading of the malaria vector Anopheles hyrcanus and the incidence of melanoma. The comparison was done with the coupled regional Eta Belgrade University and Princeton Ocean Model for the period 1961-2015 using the A1B scenario, and the expected changes up to 2030. Overall, significant correlation was found between the frequency of WNV in Culex pipiens and the overwintering temperature averages and seasonal relative humidity at the sampling sites. Correlation was also found between the spreading and relative abundance of Anopheles hyrcanus and the trend of the mean annual temperature. There was also an increase in melanoma incidence.

Minor comments to authors
Title Authors wrote "malaria vectors" but the only presented data on only one vector Anopheles hyrcanus Corrected according to the suggestion. The SRES-A1B scenario is defined in the text, and central differences to RCP explained. Due to this, authors think that selection of scenario, to some extent, is irrelevant for the presented results. The main storyline behind the A1B scenario is rapid economic growth, followed by a significant increase in greenhouse gases concentrations in the future. In the Fifth Assessment Report (AR5), the Representative Concentration Pathway (RCP) is introduced, which are possible future concentration pathways without any storyline behind them. Comparing SRES-A1B and RCPs in terms of the greenhouse gases concentrations, at the end of this century SRES-A1B is the closest to RCP6.0, but for the time horizon used in this study, up to 2030, the difference between any SRES or RCPs are relatively small.

Nomenclature of territorial units for statistics is originally abbreviated NUTS from the French version (Nomenclature des Unités territoriales statistiques).
Reviewer #2: Authors are presenting an interesting paper regarding the effects of climate change in Northern Serbia considering three independent measures: The spread of Anopheles hyrcanus, the presence of West Nile Virus in Culex pipiens, and the incidence of melanoma cases. The paper is interesting, however, discussion should be improved specially on the uncertainty of future predictions since they are using just one climatic model. Further, their results should be stated more carefully since their model rely on assumptions (e.g., manually selected variables) which are also not clearly stated.

L345-352: New text added:
The temperature trend over the period of observations used in this study and for the future time horizon following A1B scenario obtained with the EBU-POM regional climate model is within the multi-model ensemble (MME) spread of regional climate models with similar configuration used in the ENSEMBLES project (van der Linden and Mitchell, 2009). For the period 2001-2030 the temperature change for the region of interest in the EBU-POM integration is 0.75 °C concerning the period 1961-1990 and for the same period ENSEMBLES MME spread range is 0.5-1.5 °C (MEP, 2017). Following this finding, other results presented in this paper that relay on temperature change, can be seen as an estimate that will be within uncertainty related to the future temperature projection.

New references included:
van der Linden P and Mitchell JFB, editors. ENSEMBLES Climate Change and its Impacts. The paper is showing results in the order: malaria vector, WNV, and melanoma. I suggest following the same order in the abstract.

Changed as suggested.
Authors are using one of the SRES future climatic scenarios; currently the standard for future climate studies are the RCP scenarios. Authors should describe the nature of the SRES-A1B scenario, which is not mentioned in any part of the study. Further, authors should explicitly discuss uncertainty on their predictions since they are not using other scenarios or other climatic models.
Authors addressed this comment in the text corrected. Please check response to the L113-122 and L 345-352 above.
Lines 176-180. There is no discussion or results regarding these sentences. Was the comparison between EBU-POM model and the Republic Hydrometeorological Service of Serbia perfect?
What is the implication of this approach on the overall paper? This is a valuable comment since the information measure(s) is(are) a good indicator of the reliability of model outputs and thus on the overall paper. The increasing complexity of climate models is a growing concern in the modelling community. However, because we invested a serious effort to make our models more "realistic", we included more parameters and processes. With increasing model complexity, we are less able to manage and understand the model behaviour. Thus, from a user's perspective, the following question is entirely natural: "How complex model (EBU-POM model in our case) do I need to use to study this problem (assessment of climate change impact on malaria vectors, West Nile disease, and incidence of melanoma in the VPS) with this data set (temperature and/or precipitation)?". In the revised version, we inserted the additional text. Line 277-280: There is no evidence in this paper supporting this affirmation since the variables analyzed corresponded to three temperature related variables and just one considering humidity. Moreover, results were never compared statistically; modify accordingly.

L229-249: New text added:
Corrections made as suggested. The sentence "It seems that temperature in semi-urban areas dominates the other environmental factors influencing WNV circulation in nature (e.g. landscape suitability for reservoir host and mosquito vector, host availability, precipitation), as it is the primary factor affecting both mosquito vector abundance and virus replication." now reads as: L400-404: Corrected to: It seems that temperature in semi-urban areas is an essential environmental factor influencing WNV circulation (landscape suitability for reservoir host and mosquito vector, host availability, and precipitation/water availability are somewhat similar in investigated semi-urban areas of VPS), as it affects both mosquito vector abundance and virus replication.  considering some examples. We agree with V&P ideas, but we did not find the place where they explicitly say that it would always be using the incidence rate instead of cumulative (the question of overestimation). To our understanding, they left the space for a situation when the use of cumulative incidence gives acceptable results. For example, it is partly seen in the paper by Wu et al. (2014). There is another moment why we used cumulative incidence. It is well-known that there is a high correlation between sun exposure (and received cumulated doses of the UV radiation) and melanoma. If that doses (or any climate element) on a daily basis are used from regional climate models, they cannot be directly correlated with daily or monthly measured or calculated biological quantities. The reason for that is the fact that from regional climate models, we can estimate just the trend of the considered physical quantity (in our case -UV doses through their cumulative values). Correspondingly it is correlated with cumulative incidence. Having said that, after the end of the statement in Line 336, we inserted the following text.
The legend in Fig. 4(c) and y-axis in Fig 4 (d) are changed as suggested.
The M&M -Melanoma incidence and UVR was amended by the following text:

L216-221: New text added:
In the analysis we have used two indicators: (i) melanoma incidence rate that is a measure of the number of new cases ("incidence") per unit of time ("rate") and (ii) cumulative incidence ("incidence proportion" that measures the number of new cases per person in the population over a defined period of time -often called risk or proportion). Melanoma incidence rate (per 100,000 people) for ten years 1995 -2004 was based on the data obtained from the paper by Jovanović et al. [7]. From these data, we calculated the cumulative incidence.
The discussion was also amended by the following text:

L437-441: New text added:
In a cohort study, Wu et al. (2014) considered the impact of long-term UV radiation flux on skin cancer risk. Comparing with participants in the lowest quintile of cumulative UV radiation flux in adulthood, they found that participants in the highest quintile had multivariable-adjusted risks (cumulative incidence). According to Vandenbroucke and Pearce (2012), some studies where cumulative incidence is used can over-represent the trends. In table S3 consider adding the number of mosquito samples per site.

Number of samples added in the table.
Figure 6 can be replaced with the statistics of such graphic for readers' interpretation.

L442-446:
Corrections made as suggested. Figure 6 deleted, the paragraph now reads as: From a statistical point of view, the linear regression model for modelling the cumulative incidence of melanoma versus the difference of the cumulative UVR doses for hot and warm days (Fig 4d) is acceptable. Parameters are statistically highly significant (r = 0.9712 and p = 0.000003) while analysis of residual distribution shows a good agreement with the normal distribution (Shapiro-Wilk test, W = 0.9608, p = 0.7952).
Authors are justifying the paper under the 'One Health' concept, however they are not discussing the idea further. I would like discussing explicitly the benefits of putting together a set of multidisciplinary specialists to the development of the manuscript and how this contribution is part of the one health concept.
The discussion was amended by the following text:

L322-344: New text added:
Despite globalisation trends, researchers have become "closed" in their ever-smaller communication circles which are not limited by state or national borders but by the professional language and way of thinking. Thus, by the end of the 20 th century, the scientific community has been faced with problems in communication within its confines. One of the principal reasons why vector-borne diseases (VBD) are so difficult to predict, is the complex interaction of multiple factors (vector, host, pathogen, environment including short-term weather patterns and long-term climate change) in space and time ). Only groups from multiple sectors that communicate and work together on specific aspects of VBD systems will be able to answer the most exciting and pressing problems in the field (Moore 2008). Authors of this paper started collaboration in 2003 comparing the climates of the foci of WNV circulation in USA (California Central Valley) and Europe (Bucharest area) with VPS. As compared climates showed quite similar patterns, colleagues from public health and veterinary joined the initial group of meteorologists and medical entomologists. With the idea to better draw upon the resources and insights of the various sectors we designed and implemented research and programmes to achieve better outcomes in the control of zoonoses