Epidemiological Analysis of Environmental Factors Affecting Porcine Pleuropneumonia in a Herd Endemic for Actinobacillus pleuropneumoniae

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In the article “Epidemiological analysis of environmental factors affecting porcine pleuropneumonia caused by Actinobacillus pleuropneumoniae” outer climate conditions in a four year period were described and compared with mortality in finishing pigs in one farm in respective months. The topic is relevant because respiratory disease in pigs are still an important economic and welfare probelm in pig productio. 

Unfortunately the current manuscript has substantial methodological and conceptual limitations. The study design as described in the article does not allow attribution of mortality to A. pleuropneumoniae. Results of used statistical approaches are inadequately described.

The article should be rewritten in a major revision if possible.

1. The central statement of the study is, that mortality is associated with infection by Actinobacillus pleuropneumoniae, but the data  represent only mortality in finishing pigs. It should be described, if mortality is specific only to respiratory disease. How can the authors be sure, that mortality is due only to APP ?
It is not describe, if and how this diagnosis was confirmed. How many pigs in which periods were sent to necropsy (Proportion of died/euthanized pigs which were necropsied followed by subsequent diagnostical methods as bacteriological examination of the respiratory tract).
Later in the text (discussion) is stated, that several pathogens associated with the Porcine Respiratory Disease Complex can be involved in disease/mortality in this farm (line 358-368).
The evaluation of microbiological diagnostical findings in this farm should be added in a specific way (number/percentage of pathogens occurred in which month/season/period, clinical signs in different seasons (how many animals affected) or at least percentage of animals with lung lesions at slaughter, which can be related to the climatic conditions instead of mortality (causes of death are unclear so far). The manuscript frequently suggests causal relationships between climatic factors and APP outbreaks. However, the study design is only observational. It can be stated, that “climatic variables were associated with variations in mortality.”  But then also results of the statistical evaluation should be elaborated (see below).

2. All data originate only from one farm and this limitation should be addressed in the discussion. Farm-specific management practices could be responsible for the observed trends. The ventilation system should be described in detail, at best with parallel data from internal stable measurements of temperature and humidity in relation to the outer climate. The temperature-humidity-index is only mentioned in the discussion. It can be calculated for the observation periods and included in the discussion. Known critical THI should be related to the observed THI in the study. On farms changes in biosecurity, ventilation, nutrition, or treatment protocols over four years are not controlled or described in the study. The vaccination scheme should be described. Generally, the findings cannot be generalized to other production systems. The study can be considered as a case report. This limitation should be discussed

3. The meteorological data were obtained from AccuWeather regional data, but no indoor climate data from the stable are shown, although respiratory health is strongly influenced by indoor environmental conditions, including ventilation rate, ammonia levels, dust load, stocking density and especially indoor temperature gradients. It should be explained how the ventilation system in the farm regulates microclimate conditions under specific outdoor weather conditions in the different seasons.

4. The statistical approaches are not clear. They are mentioned in the material and methods section, but they do not appear again in the results section. I cannot understand the relevant Figure 6. In case, that correlation coefficients are shown in the squares, they are different to those written in the text. I cannot see significant correlations.

The dataset contains time-series structure, but only simple Pearson correlations were performed/shown. A multiple linear regression model is mentioned but results are not reporte. It is stated that a SARIMA model was applied, but model parameters are not reported and model outcomes are missing. At present, the statistical analysis is insufficiently documented and not reproducible.

Minor comments:

Line 15: “..piglet mortality..…in herds”, please rephrase, because finishing pigs are no piglets and observations were only made in one herd.

Tables 1-4 should be combined in a single summarizing table. The graphical representations of seasonal trends, should also be combined, e.g.  with 4 individual dots in one column (month), indicating in addition the mean/median as a graph.

Discussion: Instead of repeating the results, the pathomechanism of climate factors influencing infection  should be discussed. What is known ?

Comments on the Quality of English Language

Minor mistakes in grammar and spelling should be corrected

Author Response

Comment 1: 1. The central statement of the study is, that mortality is associated with infection by Actinobacillus pleuropneumoniae, but the data  represent only mortality in finishing pigs. It should be described, if mortality is specific only to respiratory disease. How can the authors be sure, that mortality is due only to APP ?
It is not describe, if and how this diagnosis was confirmed. How many pigs in which periods were sent to necropsy (Proportion of died/euthanized pigs which were necropsied followed by subsequent diagnostical methods as bacteriological examination of the respiratory tract).
Later in the text (discussion) is stated, that several pathogens associated with the Porcine Respiratory Disease Complex can be involved in disease/mortality in this farm (line 358-368).
The evaluation of microbiological diagnostical findings in this farm should be added in a specific way (number/percentage of pathogens occurred in which month/season/period, clinical signs in different seasons (how many animals affected) or at least percentage of animals with lung lesions at slaughter, which can be related to the climatic conditions instead of mortality (causes of death are unclear so far). The manuscript frequently suggests causal relationships between climatic factors and APP outbreaks. However, the study design is only observational. It can be stated, that “climatic variables were associated with variations in mortality.”  But then also results of the statistical evaluation should be elaborated (see below).
Response 1: We thank the reviewer for this important and insightful comment. We agree that, given the observational design of the study and the multifactorial nature of the Porcine Respiratory Disease Complex, it is not possible to attribute all recorded mortality exclusively to Actinobacillus pleuropneumoniae. While this pathogen was consistently identified on the farm, we acknowledge that other respiratory pathogens may have contributed to the observed mortality, and therefore a direct causal relationship cannot be firmly established. To address this concern, we have revised the manuscript to improve clarity and avoid overinterpretation. Specifically, we have replaced the expression “mortality due to APP” with “mortality due to respiratory diseases” in lines 23 and 389. In addition, statements suggesting causality between climatic factors and disease outbreaks have been reformulated, and the expression “climatic variables were associated with variations in mortality” has been introduced in lines 378–379, 392–393, and 433–434.

Regarding diagnostic confirmation, the presence of Actinobacillus pleuropneumoniae on the farm was monitored using PCR testing performed at regular intervals of approximately every 4–6 months. However, we acknowledge that these periodic assessments do not allow the attribution of each individual mortality case specifically to APP infection. Concerning post-mortem investigations, approximately 90% of deceased pigs were subjected to necropsy, which enabled the identification of lesions consistent with respiratory disease. Furthermore, in response to the reviewer’s suggestion to provide a more robust indicator of respiratory health status, we have incorporated additional considerations regarding the occurrence of lung lesions observed at slaughter and their potential association with climatic conditions. This aspect has been added and discussed in lines 443–451. These revisions were implemented to ensure a more accurate interpretation of the findings and to align the conclusions with the limitations inherent to the study design.

Comment 2: 2. All data originate only from one farm and this limitation should be addressed in the discussion. Farm-specific management practices could be responsible for the observed trends. The ventilation system should be described in detail, at best with parallel data from internal stable measurements of temperature and humidity in relation to the outer climate. The temperature-humidity-index is only mentioned in the discussion. It can be calculated for the observation periods and included in the discussion. Known critical THI should be related to the observed THI in the study. On farms changes in biosecurity, ventilation, nutrition, or treatment protocols over four years are not controlled or described in the study. The vaccination scheme should be described. Generally, the findings cannot be generalized to other production systems. The study can be considered as a case report. This limitation should be discussed.
Response 2: We fully acknowledge that the data originate from a single farm, and therefore the findings may be influenced by farm-specific management practices and cannot be directly generalized to other production systems. This limitation has now been explicitly addressed in the Discussion section, together with additional study limitations, in lines 512–526.

In response to the reviewer’s suggestion regarding environmental indicators, we have expanded the Discussion to include the temperature–humidity index (THI). Specifically, THI has now been incorporated and discussed in relation to previously reported thresholds from the literature, allowing for a more comprehensive interpretation of the climatic conditions observed during the study period (lines 483–511).

We have also improved the description of farm management and housing conditions. Details regarding the ventilation system, including relevant modifications during the study period, have been added in lines 98–110. Furthermore, information on changes in biosecurity measures, ventilation, nutrition, treatment protocols, and the vaccination scheme has been included and clarified in lines 101–135, as suggested. We appreciate the reviewer’s recommendation to consider the study as a case report. If, following the review process, this classification is deemed more appropriate by the editors, we are open to revising the manuscript accordingly.

Comment 3: 3. The meteorological data were obtained from AccuWeather regional data, but no indoor climate data from the stable are shown, although respiratory health is strongly influenced by indoor environmental conditions, including ventilation rate, ammonia levels, dust load, stocking density and especially indoor temperature gradients. It should be explained how the ventilation system in the farm regulates microclimate conditions under specific outdoor weather conditions in the different seasons.
Response 3: We would like to clarify that the farm in this study was equipped with a conventional ventilation system, which operates without advanced automated controls and is therefore directly influenced by outdoor weather conditions across different seasons. Accordingly, we have added a statement addressing this point in the manuscript (lines 98-100) to explain how the ventilation system regulates indoor microclimate conditions under varying external weather conditions.

Comment 4: The statistical approaches are not clear. They are mentioned in the material and methods section, but they do not appear again in the results section. I cannot understand the relevant Figure 6. In case, that correlation coefficients are shown in the squares, they are different to those written in the text. I cannot see significant correlations. The dataset contains time-series structure, but only simple Pearson correlations were performed/shown. A multiple linear regression model is mentioned but results are not reported. It is stated that a SARIMA model was applied, but model parameters are not reported and model outcomes are missing. At present, the statistical analysis is insufficiently documented and not reproducible.
Response 4: We thank the reviewer for the valuable feedback regarding the clarity and reporting of the statistical analyses. In response, Figure 6 has been removed from the manuscript to avoid confusion. Additionally, we have now expanded the Results section (lines 354-370) to include the outcomes of both the multiple linear regression model and the SARIMA model. These additions provide a clear link between the statistical approaches described in the Materials and Methods and the corresponding results, ensuring that the analyses are fully documented and reproducible.

Comment 5: Minor Line 15: “..piglet mortality..…in herds”, please rephrase, because finishing pigs are no piglets and observations were only made in one herd. Tables 1-4 should be combined in a single summarizing table. The graphical representations of seasonal trends, should also be combined, e.g.  with 4 individual dots in one column (month), indicating in addition the mean/median as a graph. Discussion: Instead of repeating the results, the pathomechanism of climate factors influencing infection  should be discussed. What is known ?
Response 5: We have revised the manuscript accordingly. The wording in Lines 15-16 has been corrected to accurately reflect that the study refers to finishing pigs from a single herd. In response to the comment regarding data presentation, Tables 1–4 have been consolidated into a single comprehensive table to improve readability and facilitate interpretation of the results. Additionally, the graphical representations of seasonal trends have been combined into a single figure, as suggested, allowing for a clearer visualization of monthly variations and overall trends; this revised figure has been included in lines 316–317.

Furthermore, the Discussion section has been strengthened by adding a dedicated paragraph addressing the pathomechanisms through which climatic factors may influence respiratory infections, based on current knowledge from the literature. This addition can be found in lines 473–482.

Reviewer 2 Report

Comments and Suggestions for Authors

The data for this four-year observational study is from a single commercial farm in the Republic of Moldova. The data measures the relationship between weather conditions, temperature, humidity, and rainfall, and the mortality rate of piglets and fattening pigs in a continuously endemically infected herd with Actinobacillus pleuropneumoniae. This is a very important topic in the swine industry, and teasing out the relationship between environmental stressors and disease outcome is a very important aspect in disease control. A very strong aspect of the study is the long-term data, 48 months, from a single farm. The data has a clear seasonal pattern, with mortality rates increasing in spring and autumn and decreasing in summer and winter. The data is very clear, and the main result is simple. However, the study has some very serious methodological and diagnostic flaws.

Major Comments

Critical Flaw: Lack of definitive APP diagnosis:

This is the most significant issue. The manuscript attributes mortality directly to APP based on clinical signs and the farm's "endemic" status. However, no diagnostic data (bacteriology, PCR, serology, or even detailed post-mortem lesion scoring specific to APP) are provided to confirm APP as the primary cause of death during the study period.

The clinical signs described (lethargy, dyspnea, fever) and lesions mentioned (hemorrhagic necrotizing pneumonia, fibrinous pleuritis) are highly suggestive of APP but are not pathognomonic. Other pathogens in the Porcine Respiratory Disease Complex (PRDC), such as Pasteurella multocida, Streptococcus suis, or viral agents like PRRSV and Swine Influenza Virus, can cause similar clinical presentations and lung lesions. The authors themselves acknowledge the presence of PRDC in the discussion, which introduces a major confounding variable.

Recommendation: The authors must provide evidence that APP was the primary etiology. If retrospective samples (e.g., fixed tissues, frozen bacteria) or records are available, confirmatory testing (e.g., PCR on lung tissue, serotyping of isolates) is essential. If not, the manuscript's title, abstract, and conclusions must be fundamentally revised to reflect that mortality was associated with "respiratory disease" in an APP-endemic herd, rather than being definitively caused by APP. The correlation would then be between climate and overall respiratory mortality, not specifically APP.

Confounding Variables and Study Design:

The study is an uncontrolled observational analysis from a single farm. Many critical confounders are not addressed:

Management Practices: Were there any changes in vaccination protocols, antibiotic use (prophylactic/metaphylactic), feed composition, or stocking density over the four years? Any of these could dramatically influence mortality rates independently of climate.

Co-infections: The farm is described as having a "persistent issue with respiratory diseases," implying the presence of other pathogens. The role of PRRSV, influenza, Mycoplasma hyopneumoniae, etc., in the observed mortality is unknown. Were pigs tested for these? Seasonal patterns in viral co-infections could easily confound the climate-mortality association.

Housing and Ventilation: The authors mention the "protective role of stable housing" in winter. However, no data on the indoor microclimate (temperature, humidity, ammonia levels, ventilation rates) are provided. The external weather data is a proxy, but the indoor environment is what the pigs actually experience. The effectiveness of ventilation control during transitional seasons (spring/autumn) is a critical variable that is missing.

Recommendation: The discussion must be significantly expanded to address these limitations. The authors should state clearly that while a correlation was found, causality cannot be proven due to these unmeasured confounders. Ideally, data on farm management stability and the health status regarding other major pathogens should be included.

Statistical Analysis and Interpretation:

The Pearson correlation analysis presented at the end (Line 284) is a step in the right direction but is insufficient. Aggregating four years of data into a single correlation ignores the time-series nature of the data and potential autocorrelation.

Recommendation: A more robust statistical approach is needed. For example:

Time-series analysis: Methods like ARIMA (Autoregressive Integrated Moving Average) models could better account for temporal trends and seasonality.

Multivariate regression: A model with mortality as the outcome and temperature, humidity, and precipitation as predictors, while controlling for potential confounders (if data were available) and testing for interactions, would be far more powerful. This could also identify lag effects (e.g., does a cold, humid week predict mortality 2 weeks later?).

The correlation matrix (Figure 6) shows r=0.63 for temperature and r=0.54 for humidity. While these are presented as positive correlations, the seasonal data (Tables 1-4) show that high mortality occurs in spring/autumn with moderate temperatures, while low mortality occurs in summer with high temperatures. This suggests a non-linear, U-shaped relationship, which a simple Pearson correlation fails to capture. The interpretation is therefore misleading.

Presentation of Data in Tables:

Tables 1-4: The mortality percentages are inconsistent in all tables. For example, in Table 1, March mortality is 7.27%, but the text (Line 134) says it peaked at 7.27% in March. However, Table 1 also shows April at 7.41%, which is higher. Please ensure all values in the text and tables are accurate and consistent.

Minor 

Abstract: The conclusion "indicating that climatic stress may significantly influence the susceptibility to infection" is too strong based on the data. It should be toned down to something like "...is associated with increased mortality in herds with endemic APP."

Discussion (Lines 358-368): The discussion of PRDC and co-infections is important but appears almost as an afterthought. It needs to be integrated more thoroughly. For instance, the high mortality in spring could be due to the wean-to-finish flow, where pigs are losing maternal immunity and are more susceptible to all respiratory pathogens, and this vulnerability is then exacerbated by climatic stress.

Specific technical questions for Authors

1. Can you provide any laboratory confirmation (e.g., bacterial culture, PCR results from lung tissue) that APP was the primary cause of death during this 4-year period?

2. What was the health status of the herd regarding other major respiratory pathogens, particularly PRRSV and Mycoplasma hyopneumoniae?

3. Were there any changes in management, vaccination, or antibiotic use on the farm between 2021 and 2024 that could have influenced mortality rates?

4. Do you have any data on the indoor environmental conditions (temperature, humidity) within the fattening units to compare with the external weather data?

 

Author Response

Comment 1: Critical Flaw: Lack of definitive APP diagnosis:

This is the most significant issue. The manuscript attributes mortality directly to APP based on clinical signs and the farm's "endemic" status. However, no diagnostic data (bacteriology, PCR, serology, or even detailed post-mortem lesion scoring specific to APP) are provided to confirm APP as the primary cause of death during the study period.

The clinical signs described (lethargy, dyspnea, fever) and lesions mentioned (hemorrhagic necrotizing pneumonia, fibrinous pleuritis) are highly suggestive of APP but are not pathognomonic. Other pathogens in the Porcine Respiratory Disease Complex (PRDC), such as Pasteurella multocida, Streptococcus suis, or viral agents like PRRSV and Swine Influenza Virus, can cause similar clinical presentations and lung lesions. The authors themselves acknowledge the presence of PRDC in the discussion, which introduces a major confounding variable.

Recommendation: The authors must provide evidence that APP was the primary etiology. If retrospective samples (e.g., fixed tissues, frozen bacteria) or records are available, confirmatory testing (e.g., PCR on lung tissue, serotyping of isolates) is essential. If not, the manuscript's title, abstract, and conclusions must be fundamentally revised to reflect that mortality was associated with "respiratory disease" in an APP-endemic herd, rather than being definitively caused by APP. The correlation would then be between climate and overall respiratory mortality, not specifically APP.

Response 1: We fully agree that, based on the available data, it is not possible to definitively attribute all mortality events exclusively to APP, particularly in the context of the multifactorial etiology of the Porcine Respiratory Disease Complex.

To address this concern, we have clarified the diagnostic approach used on the farm. PCR testing for Actinobacillus pleuropneumoniae was performed at regular intervals of approximately every 4–6 months, confirming the presence and persistence of the pathogen within the herd over time. These aspects have now been explicitly described in the revised manuscript (lines 444–453). However, we acknowledge that these periodic diagnostic evaluations do not allow for the attribution of each individual mortality case specifically to APP infection, nor do they exclude the contribution of other bacterial or viral pathogens commonly involved in respiratory disease in swine.

In light of this limitation, and in agreement with the reviewer’s recommendation, we have revised the manuscript to avoid overstatement of causality. Specifically, the terminology has been adjusted to reflect “mortality due to respiratory diseases” rather than direct attribution to APP, including in lines 23 and 389.

Comment 2: Confounding Variables and Study Design:

The study is an uncontrolled observational analysis from a single farm. Many critical confounders are not addressed:

Management Practices: Were there any changes in vaccination protocols, antibiotic use (prophylactic/metaphylactic), feed composition, or stocking density over the four years? Any of these could dramatically influence mortality rates independently of climate.

Co-infections: The farm is described as having a "persistent issue with respiratory diseases," implying the presence of other pathogens. The role of PRRSV, influenza, Mycoplasma hyopneumoniae, etc., in the observed mortality is unknown. Were pigs tested for these? Seasonal patterns in viral co-infections could easily confound the climate-mortality association.

Housing and Ventilation: The authors mention the "protective role of stable housing" in winter. However, no data on the indoor microclimate (temperature, humidity, ammonia levels, ventilation rates) are provided. The external weather data is a proxy, but the indoor environment is what the pigs actually experience. The effectiveness of ventilation control during transitional seasons (spring/autumn) is a critical variable that is missing.

Recommendation: The discussion must be significantly expanded to address these limitations. The authors should state clearly that while a correlation was found, causality cannot be proven due to these unmeasured confounders. Ideally, data on farm management stability and the health status regarding other major pathogens should be included.
Response 2: We have substantially expanded the Discussion section to explicitly address these limitations (lines 513–527). Specifically, we now clarify that all data were derived from a single commercial farm, which may limit the generalizability of the findings to other production systems with different management practices, housing conditions, or health status. We also state that, although mortality was analyzed in the context of respiratory disease dynamics, confirmatory diagnostic testing was not systematically performed for all affected animals, which restricts the ability to attribute outcomes to specific etiological agents; therefore, mortality is more appropriately interpreted as being associated with respiratory conditions rather than a single pathogen. In addition, we highlight that the study relied on external meteorological data, while detailed information on the internal microclimate of the facilities (e.g., temperature, humidity, ventilation efficiency) was not available. Furthermore, additional information regarding farm management practices has been included in lines 99–138, covering aspects such as vaccination protocols, antimicrobial use, nutrition, and housing conditions, in order to provide better context for interpretation.

Comment 3: Statistical Analysis and Interpretation:

The Pearson correlation analysis presented at the end (Line 284) is a step in the right direction but is insufficient. Aggregating four years of data into a single correlation ignores the time-series nature of the data and potential autocorrelation.

Recommendation: A more robust statistical approach is needed. For example:

Time-series analysis: Methods like ARIMA (Autoregressive Integrated Moving Average) models could better account for temporal trends and seasonality.

Multivariate regression: A model with mortality as the outcome and temperature, humidity, and precipitation as predictors, while controlling for potential confounders (if data were available) and testing for interactions, would be far more powerful. This could also identify lag effects (e.g., does a cold, humid week predict mortality 2 weeks later?).

The correlation matrix (Figure 6) shows r=0.63 for temperature and r=0.54 for humidity. While these are presented as positive correlations, the seasonal data (Tables 1-4) show that high mortality occurs in spring/autumn with moderate temperatures, while low mortality occurs in summer with high temperatures. This suggests a non-linear, U-shaped relationship, which a simple Pearson correlation fails to capture. The interpretation is therefore misleading.

Presentation of Data in Tables:

Tables 1-4: The mortality percentages are inconsistent in all tables. For example, in Table 1, March mortality is 7.27%, but the text (Line 134) says it peaked at 7.27% in March. However, Table 1 also shows April at 7.41%, which is higher. Please ensure all values in the text and tables are accurate and consistent.
Response 3: We thank the reviewer for this detailed and constructive comment regarding the statistical analysis and data interpretation. In response to similar concerns raised by another reviewer, we have removed Figure 6, which contained the Pearson correlation matrix. To address the limitations of the initial statistical approach, we have substantially strengthened the analysis by introducing both a multiple linear regression model and a SARIMA time-series model, which are now presented in the Results section (lines 354–371). In addition, we carefully reviewed the entire manuscript to ensure consistency between tables and text. Several discrepancies in reported values have been corrected accordingly in lines 180–183, 187, 190, 199, 222–224, 229, 231, 258, 266, 296 and 301.

Comment 4: Minor 

Abstract: The conclusion "indicating that climatic stress may significantly influence the susceptibility to infection" is too strong based on the data. It should be toned down to something like "...is associated with increased mortality in herds with endemic APP."

Discussion (Lines 358-368): The discussion of PRDC and co-infections is important but appears almost as an afterthought. It needs to be integrated more thoroughly. For instance, the high mortality in spring could be due to the wean-to-finish flow, where pigs are losing maternal immunity and are more susceptible to all respiratory pathogens, and this vulnerability is then exacerbated by climatic stress.

Specific technical questions for Authors

1. Can you provide any laboratory confirmation (e.g., bacterial culture, PCR results from lung tissue) that APP was the primary cause of death during this 4-year period?

2. What was the health status of the herd regarding other major respiratory pathogens, particularly PRRSV and Mycoplasma hyopneumoniae?

3. Were there any changes in management, vaccination, or antibiotic use on the farm between 2021 and 2024 that could have influenced mortality rates?

4. Do you have any data on the indoor environmental conditions (temperature, humidity) within the fattening units to compare with the external weather data?

Response 4: We agree that the original phrasing in the abstract was too strong relative to the data presented. Therefore, we have revised the sentence to better reflect the associative nature of our findings. The abstract has been modified according to the reviewer’s suggestion in lines 21–24, where the conclusion now states that climatic stress is associated with increased mortality in herds with endemic APP, rather than implying a direct causal relationship.
We appreciate the reviewer’s suggestion to better integrate the discussion of Porcine Respiratory Disease Complex and co-infections. In response, we expanded this section to more thoroughly address the multifactorial nature of respiratory disease in intensive pig production systems. Additional discussion regarding the potential role of co-infections and the interaction between seasonal physiological vulnerability and climatic stress has been incorporated in lines 465–480 of the revised manuscript.
Regarding laboratory confirmation, routine molecular monitoring for APP was performed on the farm. PCR testing was conducted every 4–6 months during the study period to monitor the presence of APP in the herd, supporting its endemic status.
Information regarding changes in therapy protocols, management practices, and vaccination schemes applied on the farm between 2021 and 2024 has been detailed in lines 100–139 of the revised manuscript.
We acknowledge the reviewer’s suggestion to compare external weather data with indoor environmental parameters. However, data regarding indoor environmental conditions (temperature and humidity) within the fattening units were not available for the study period, and therefore such comparisons could not be performed. This limitation has been acknowledged in the manuscript (lines 521-527).

Reviewer 3 Report

Comments and Suggestions for Authors

This study analyzed four-year farm data (2021–2024) to examine associations between climatic factors (temperature, humidity, precipitation) and mortality from Actinobacillus pleuropneumoniae in finishing pigs, showing seasonal peaks in spring and autumn linked to environmental stress.

1.Data were derived from a single commercial farm, limiting external validity and preventing generalization to other production systems or geographic regions.

2.The analysis relied on observational retrospective farm records, which may contain inconsistencies or reporting bias.

3.APP infection was inferred from farm disease history and mortality patterns, while microbiological confirmation, pathogen load, or serotyping data were not systematically reported. Fourth, several potential confounders were not controlled, including stocking density, ventilation performance, management practices, feed changes, and co-infections within the porcine respiratory disease complex (PRDC).

 4.The statistical analysis primarily used correlation and regression, which cannot establish causal relationships between climatic variables and mortality.

5.Meteorological data were obtained from regional weather databases rather than on-farm microclimate measurements, which may not accurately represent actual barn conditions affecting pigs.

6.Redundant explanations in Results and Discussion;Occasional overinterpretation of correlations

Suggestion:Clarify the study design and explicitly state its observational nature. Strengthen the Methods by describing data sources, farm management conditions, and potential confounders (e.g., ventilation, stocking density, vaccination). Add a clearer statistical explanation and justify the use of correlations. In the Discussion, compare findings with more international studies and moderate causal claims. Expand the Limitations section to acknowledge the single-farm dataset, lack of pathogen confirmation, and possible effects of co-infections and management factors.

7. Language issues:

“Episodes” and “losses” are imprecise.

Grammatical error: with an additionally reporting serotypes 2 and 6.

“Prevalence” and “incidence” are mentioned but not clearly defined in the analysis.Clarify or remove if not analyzed. Suggestion:The epidemiological parameters collected included mortality and monthly incidence.

Author Response

Comment 1:1.Data were derived from a single commercial farm, limiting external validity and preventing generalization to other production systems or geographic regions. 2.The analysis relied on observational retrospective farm records, which may contain inconsistencies or reporting bias.
Response 1: Thank you for this important observation. We acknowledge that the use of data derived from a single commercial farm may limit the external validity of the findings and restrict their generalizability to other production systems or geographic regions. This aspect has been explicitly addressed in the revised manuscript, where we have included it as a study limitation in the Discussion section (Lines 513–527). There, we emphasize that the results should be interpreted within the context of the specific farm conditions and highlight the need for further studies across multiple farms and regions to validate and extend these findings.

Comment 2: 3.APP infection was inferred from farm disease history and mortality patterns, while microbiological confirmation, pathogen load, or serotyping data were not systematically reported. Fourth, several potential confounders were not controlled, including stocking density, ventilation performance, management practices, feed changes, and co-infections within the porcine respiratory disease complex (PRDC).
Response 2: In the revised manuscript, we have clarified that PCR testing for APP was performed at regular intervals of 4–6 months and this information has now been explicitly included in the Materials and Methods section (lines 136–139) and further referenced in the Discussion (lines 444–453) to support the interpretation of APP as an endemic pathogen in the herd during the study period. In addition, to address the reviewer’s concern regarding potential confounding factors, we have expanded the description of farm-level variables. Information regarding farm management, including ventilation systems, nutrition, therapeutic protocols, and vaccination schemes applied during the study period, has been added in lines 98–139 of the revised manuscript. 

Comment 3: The statistical analysis primarily used correlation and regression, which cannot establish causal relationships between climatic variables and mortality.
Response 3: To address this limitation and provide a more robust assessment of temporal patterns, we have expanded the statistical analysis by incorporating both a SARIMA (Seasonal Autoregressive Integrated Moving Average) model and a multiple regression model (Lines 354-371), which allow for the evaluation of seasonality, temporal dependencies, and multivariate effects, while still recognizing that causality cannot be conclusively inferred from observational data.

Comment 4: 5.Meteorological data were obtained from regional weather databases rather than on-farm microclimate measurements, which may not accurately represent actual barn conditions affecting pigs. 6.Redundant explanations in Results and Discussion; Occasional overinterpretation of correlations Suggestion: Clarify the study design and explicitly state its observational nature. Strengthen the Methods by describing data sources, farm management conditions, and potential confounders (e.g., ventilation, stocking density, vaccination). Add a clearer statistical explanation and justify the use of correlations. In the Discussion, compare findings with more international studies and moderate causal claims. Expand the Limitations section to acknowledge the single-farm dataset, lack of pathogen confirmation, and possible effects of co-infections and management factors.
Response 4: We have carefully revised the manuscript to address these points. Specifically, we have maintained a clear acknowledgment of the study limitations in the Discussion section, including the absence of on-farm microclimate measurements and the reliance on regional meteorological data, which may not fully reflect actual barn conditions affecting pigs (Lines 513-527).

Comment 5: Language issues: “Episodes” and “losses” are imprecise. Grammatical error: with an additionally reporting serotypes 2 and 6. “Prevalence” and “incidence” are mentioned but not clearly defined in the analysis. Clarify or remove if not analyzed. Suggestion:The epidemiological parameters collected included mortality and monthly incidence.
Response 5: We have revised the manuscript to address the language and clarity issues. Specifically, the terms “episodes” and “losses” have been replaced with more precise expressions throughout the text (Lines 19–24, 169, 196, 234). Furthermore, references to “prevalence” and “incidence” have been clarified: the epidemiological parameters collected and analyzed in this study now explicitly include mortality and monthly incidence, while terms  that were not directly analyzed have been removed (line 63, 143-144, 153, 215) to avoid confusion. The phrasing in line 40 has been corrected in the revised manuscript to improve clarity and grammatical accuracy. In addition, we carefully revised the manuscript to address similar wording and grammatical issues in several other sections. Corrections were made in lines 187, 285–287, 292–293, 296–297, 379–380, 390, and 393–394 to improve readability and ensure consistent language throughout the text.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

My comments were adressed and the manuscript improved. Unfortunately I cannot find the text in alignment to the authors answers (the lines they indicate (where they made changes) are not the lines in the manuscript. I was also unable to read the comments in the pdf.

Due to the fact, that so many pigs were necropsied, at least estimates of percentages should be added.

Line 237-238 2Diagnostic monitoring was performed regularly, with polymerase chain reaction”, from which samples ? App-detection rates ?

Line 317-319 and 840-842: In case, that 90% of the died pigs were necropsied, approx. 1440 necropsies were performed per year. It would be good to give numbers on the percentages of death due to respiratory disease, detection rates of APP-like lesions,  detection rates of App and other pathogens. Mention the APP serotype. Add the results of the antimicrobial resistancy testing. How often was it performed ? Did it change over time (antimicrobial resistancy pattern ). Some more information about diagnostic would be good.This information (estimation of percentages ?) should be allocated to the different years and at best also periods in 3.1. to 3.4. I cannot find information on the slaughter lung assessment. In line 443-451 is another text.

Line 688-689 and 745, 748  “…45%  the variation in piglet mortality can be explained by the combination of the three meteorological variables” It is not “piglet” mortalitiy, but mortality in finishing pigs. Replace “piglets” by “fattening pig” or “pig” all over the text

Line 875-877: “It has been suggested that climatic factors may play a more important role in the spread of less contagious respiratory diseases, such as those caused by APP, compared to more contagious pathogens like Mycoplasma hyopneumoniae.” I do not agree, that M. hyopneumoniae is more contagious. Can you give references for that

The THI calculated in this study should be added to the results sections in the different years or in an overall statement. so far THI is addede in the discussion (which is interesting9, but the reader is not able to comper the referecne THi in the discussion with those from the study.

 

Author Response

Comment 1: Line 237-238 2Diagnostic monitoring was performed regularly, with polymerase chain reaction”, from which samples ? App-detection rates ?
Response 1: . In the revised manuscript, we have specified the types of samples used for PCR-based diagnostic monitoring in lines 139–142 and 409–410. In addition, information regarding APP detection rates has now been included in lines 410–411 and 490.

Comment 2: Line 317-319 and 840-842: In case, that 90% of the died pigs were necropsied, approx. 1440 necropsies were performed per year. It would be good to give numbers on the percentages of death due to respiratory disease, detection rates of APP-like lesions,  detection rates of App and other pathogens. Mention the APP serotype. Add the results of the antimicrobial resistancy testing. How often was it performed ? Did it change over time (antimicrobial resistancy pattern ). Some more information about diagnostic would be good.This information (estimation of percentages ?) should be allocated to the different years and at best also periods in 3.1. to 3.4. I cannot find information on the slaughter lung assessment. In line 443-451 is another text.
Response 2: We have included approximate percentage data available for the entire study period, which are now reported in lines 144–145, 404–409, and 495–500. Regarding antimicrobial resistance, relevant information has been added in lines 490–491. We would like to clarify that data were available exclusively from necropsy examinations, and not from slaughter lung assessments. This limitation has been addressed in the revised manuscript. However, based on necropsy findings, we have now included in lines 497–502 the percentages of lesions compatible with APP infection as well as other pathological conditions, to better support the interpretation of respiratory disease involvement. We would like to note that information regarding the APP serotype has been added in line 411 of the revised manuscript.

Comment 3: Line 688-689 and 745, 748  “…45%  the variation in piglet mortality can be explained by the combination of the three meteorological variables” It is not “piglet” mortalitiy, but mortality in finishing pigs. Replace “piglets” by “fattening pig” or “pig” all over the text
Response 3: We agree that the use of the term “piglet” was inaccurate in this context. Accordingly, we have corrected the terminology throughout the manuscript, replacing “piglet” with “pigs” to accurately reflect that the data refers to finishing animals. These corrections have been made in lines 19, 107, 119, 132, 362, 366, and 369.

Comment 4: Line 875-877: “It has been suggested that climatic factors may play a more important role in the spread of less contagious respiratory diseases, such as those caused by APP, compared to more contagious pathogens like Mycoplasma hyopneumoniae.” I do not agree, that M. hyopneumoniae is more contagious. Can you give references for that
Response 4: We agree that the comparison regarding the contagiousness of Mycoplasma hyopneumoniae may be debatable and insufficiently supported. In response, we have revised the phrasing in line 530, replacing “...compared to more contagious pathogens like Mycoplasma hyopneumoniae” with a more neutral expression referring to “other contagious pathogens.”

Comment 5:The THI calculated in this study should be added to the results sections in the different years or in an overall statement. so far THI is addede in the discussion (which is interesting9, but the reader is not able to comper the referecne THi in the discussion with those from the study.
Response 5: We thank the reviewer for this helpful suggestion. We agree that presenting the Temperature–Humidity Index (THI) data in the Results section improves clarity and allows for a more direct comparison with the values discussed later in the manuscript. Accordingly, we have included the THI data in the Results section in lines 379–403, providing a clearer presentation of the findings.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have made substantial efforts to address the major concerns raised in the previous review.

Title still specifies "Caused by Actinobacillus pleuropneumoniae": Given the acknowledged diagnostic limitations, the title overstates certainty. Consider changing to "Epidemiological Analysis of Environmental Factors Affecting Porcine Pleuropneumonia in a Herd Endemic for Actinobacillus pleuropneumoniae" or similar. However, this is not a fatal flaw as the text clarifies the association.

Author Response

Comment 1: Title still specifies "Caused by Actinobacillus pleuropneumoniae": Given the acknowledged diagnostic limitations, the title overstates certainty. Consider changing to "Epidemiological Analysis of Environmental Factors Affecting Porcine Pleuropneumonia in a Herd Endemic for Actinobacillus pleuropneumoniae" or similar. However, this is not a fatal flaw as the text clarifies the association.
Response 1: We thank the reviewer for this helpful suggestion. In response, we have revised the manuscript title to better reflect the epidemiological nature of the study and the diagnostic limitations acknowledged in the manuscript. The title has been modified accordingly in the revised version.

Reviewer 3 Report

Comments and Suggestions for Authors

Authors addressed all the comments, I have no further comments.

Author Response

Comment 1: Authors addressed all the comments, I have no further comments.

Response 1: We thank the reviewer for the time and effort dedicated to evaluating our manuscript and for acknowledging the revisions made in response to the previous comments. We appreciate the reviewer’s positive feedback.