Chlorophyll Fluorescence Imaging Combined with Active Oxygen Metabolism for Classification of Similar Diseases in Cucumber Plants
Round 1
Reviewer 1 Report
The study focuses on the identification of different cucumber diseases at different stages. There are comments that can contribute to improving the manuscript.
More numerical data should be added to the Abstract.
Please indicate the novelty of the study in the Abstract.
line 95: Why only one type of cucumber was used in the experiments?
line 99: Please indicate whether the analysis was carried out for both years together or for each year separately.
line 102: 130 plants for each year?
2.2 Pathogen preparation and plant inoculation: Was it not possible to obtain naturally infected samples?
line 228: 340 plant samples or 340 images?
lines 236-238: It should be deleted.
Why does Table 2 not include a statistical comparison of means?
Discriminant results could include also other classification performance metrics, such as errors, precision, F-Measure, ROC (Receiver Operating Characteristic) Area, and PRC (Precision-Recall) Area in addition to accuracy.
Directions for future research should be discussed in detail and indicated in the Conclusions.
English changes are required.
Author Response
The study focuses on the identification of different cucumber diseases at different stages. There are comments that can contribute to improving the manuscript.
More numerical data should be added to the Abstract.
Response: Thanks for your suggestion. We have added more numerical data in the abstract. (Line 19-22)
Please indicate the novelty of the study in the Abstract.
Response: Thanks for your suggestion. We have added more information in the abstract. (Line 12-13)
line 95: Why only one type of cucumber was used in the experiments?
Response: The purpose of this paper is to explore the differences in the response of chlorophyll fluorescence parameters between two diseases, while, different varieties of cucumber plants may cause differences in their growth and their responses to diseases, which will cause differences in chlorophyll fluorescence parameters. Therefore, in order to avoid interference from other factors, the same variety is generally used for experiments.
line 99: Please indicate whether the analysis was carried out for both years together or for each year separately.
Response: Thanks for your suggestion. The analysis was carried out for both years together. We have modified in Line 102-103.
line 102: 130 plants for each year?
Response: We apologize for the error about the sample number. Actually, there were total 135 plants form both years, as described in section 2.2, 50 plants for C. cassiicola group, 50 plants for C. orbiculare group and 35 from control group. We have modified in Line 102-103.
2.2 Pathogen preparation and plant inoculation: Was it not possible to obtain naturally infected samples?
Response: Actually, it is possible to get samples of natural infections, but when we see the natural disease spot by the naked eye, the disease degree of the sample has reached the late stage, so it is hard to collect natural infection samples for early stage. Therefore, the naturally infected samples are not suitable for this research.
line 228: 340 plant samples or 340 images?
Response: Thank you for your careful review. 340 sets of images to be exact, we have modified them in the manuscript.
lines 236-238: It should be deleted.
Response: Thanks. We have deleted it.
Why does Table 2 not include a statistical comparison of means?
Response: We appreciate the reviewer’s suggestion. Values in the same row have been analyzed to compare the differences at different times of inoculation, and values with different letters in the same row are significantly different at the 95% confidence interval.
Discriminant results could include also other classification performance metrics, such as errors, precision, F-Measure, ROC (Receiver Operating Characteristic) Area, and PRC (Precision-Recall) Area in addition to accuracy.
Response: Thanks. We agree with the reviewer’s comments, other performance metrics can show discriminant results, while, classification accuracies are the most intuitive and important results. If other performance metric is added, the table will become lengthy. Therefore, we chose the classification accuracies as the final results.
Directions for future research should be discussed in detail and indicated in the Conclusions.
Response: Thanks for your suggestion. We have added more information in the conclusion. (Line 469-475)
English changes are required.
Response: Thanks. We have revised the whole manuscript.
Reviewer 2 Report
This manuscript, "Chlorophyll Fluorescence Imaging Combined with Active Oxygen Metabolism for Classification of Similar Diseases in Cucumber Plants" (agronomy-2222260), proposes a method for early identification of two cucumber diseases: brown spot and anthracnose, using chlorophyll fluorescence imaging and active oxygen metabolism analysis. The support vector machine (SVM) and XGBoost algorithms were used to classify the diseases, with XGBoost being more effective with a classification accuracy of over 90%. The study links fluorescence parameters with active oxygen metabolism and demonstrates the potential of the proposed method for detecting different types and degrees of cucumber diseases. This research is very interesting to agronomy readers. The introduction, materials and methods, results, discussion, and conclusion of the paper are well-written and present the results in a logical and coherent manner.
I have some questions, however:
Why were only the SVM and XGBoost algorithms tested? Are there other AI algorithms with greater potential?
What is the potential of chlorophyll fluorescence imaging combined with active oxygen metabolism analysis for detecting different types and degrees of cucumber diseases?
What is the significance of this research for plant pathology programs and understanding of cucumber diseases? Can other plants be analysed with this method?
My minor suggestions include:
Alphabetize the keywords.
Standardize the nomenclature of equipment/reagents/software by including the manufacturer, city, state (using three-letter codes), and country when necessary. Check the manuscript for consistency.
Consider rewriting the questions on lines 36-38.
Change "chorophyll a present inside chloroplast" to "chlorophyll a present inside chloroplasts."
Change "not are exclusively" to "not exclusively."
Consider revising the language in lines 52-83.
Best regards.
Author Response
Why were only the SVM and XGBoost algorithms tested? Are there other AI algorithms with greater potential?
Response: Thanks. As described in section 2.5, the SVM and XGBoost both belong to machine learning algorithms. SVM is a more commonly used and traditional method for disease detection of agricultural product. However, XGBoost is a relatively new algorithm, which is rarely studied in disease detection of agricultural product. Therefore, we compared these two machine learning algorithms. There should be some other algorithms with more advantages for disease detection, such as deep learning, but these algorithms require a large amount of data, so in this research, we did not try them.
What is the potential of chlorophyll fluorescence imaging combined with active oxygen metabolism analysis for detecting different types and degrees of cucumber diseases?
Response: Actually, to be precise, active oxygen metabolism analysis can help explain why chlorophyll fluorescence parameters can distinguish between different species of early diseases. The diseased cucumber had a higher level of reactive oxygen species (ROS) accumulation than the healthy cucumbers, and the activity levels of ROS scavenging enzymes (SOD, APX, CAT and POD) of anthracnose were higher than those of brown spot. The classification results were consistent with the active oxygen metabolism analysis, the cucumber had a weaker resistance to brown spot than anthracnose, therefore at the same infection time, and the brown spot group had a more obvious diseased symptom, leading to a slightly higher misclassification rate for the anthracnose than that of the brown spot. Related descriptions were in Line 13-16, 379-383, 461-463.
What is the significance of this research for plant pathology programs and understanding of cucumber diseases? Can other plants be analysed with this method?
Response: This study provides a detection method for distinguishing plant similar diseases, and explains the internal reasons for distinguishing plant diseases from the physiological aspects. Theoretically, the chlorophyll fluorescence imaging method can be extended to other plants. Since plant disease damage can directly affect its photosynthesis, therefore, different chlorophyll fluorescence parameters can diagnose the plant's physiological states. Related descriptions were in Line 452-455.
My minor suggestions include:
Alphabetize the keywords.
Response: Thanks for your suggestion. We have modified them.
Standardize the nomenclature of equipment/reagents/software by including the manufacturer, city, state (using three-letter codes), and country when necessary. Check the manuscript for consistency.
Response: Thanks. We have modified them, and added the production type, but the equipment from China that use province instead of city and state.
Consider rewriting the questions on lines 36-38.
Response: Thanks for your suggestion. We have rewritten this sentence. (Line 39-41)
Change "chlorophyll a present inside chloroplast" to "chlorophyll a present inside chloroplasts."
Response: Sorry, we do not find "chlorophyll a present inside chloroplast" in this manuscript.
Change "not are exclusively" to "not exclusively."
Response: Sorry, we do not find "not are exclusively" in this manuscript.
Consider revising the language in lines 52-83.
Response: Thanks. We have revised this part.
Round 2
Reviewer 1 Report
The manuscript has been improved
