A Study on the Distribution Pattern of Banana Blood Disease (BBD) and Fusarium Wilt Using Multispectral Aerial Photos and a Handheld Spectrometer in Subang, Indonesia

Round 1

Reviewer 1 Report

This manuscript deals with the distribution pattern of banana blood diseases (BDB) and Fusarium wilt of banana (FWB) using multispectral aerial photo and handheld spectrometer in Subang-Indonesia. Both FWB and BDB are two significant diseases infecting banana production in this region. It is a good idea for using advanced geospatial information obtained from multispectral aerial photographs taken using an unmanned aerial vehicle (UAV), combined with the reliable field data to monitor the incidence of these two diseases. However, the key question is what are the incidences of two diseases in your surveyed field, how reliable of this methodology if you look the Figure 8 which was indicated there was no much healthy plants left. Furthermore, this manuscript is still needed for sufficient revision. Some suggestions and comments indicated below for author’s further improvement.

1.     In line 36, change keyword “Blood Diseases Banana” into “Banana Diseases Blood”.

2.     Give full name of Fusarium oxysporum f. sp. cubense. Also indicate if infected banana field presented in this manuscript is caused by Race 1 or Tropical Race 4 (TR4).

3.     In line 42, remove 1996.

4.     In line 88, correct “Zea may” into “Zea mays“.

5.     In line 159-160, give genome composition of four banana cultivars if it is possible.

6.     Add additional Figure to present typical symptom photos of FWB and BDB in individual plant; pseudostem, peduncle and other organs to show what are the main differences between these two main diseases.

7.     Give your full description how can you distinguish FWB and BDB.

8.     Linking soil pH to incidence of Fusarium wilt in line 358-359 should be placed in the discussion section.

9.     This is the same as above to apply for lines 431-435, 440. They should be placed in the discussion section.

10. Give the full name of MCARI when it is firstly appeared in the manuscript.

Author Response

Comments and Suggestions for Authors

This manuscript deals with the distribution pattern of banana blood diseases (BDB) and Fusarium wilt of banana (FWB) using multispectral aerial photo and handheld spectrometer in Subang-Indonesia. Both FWB and BDB are two significant diseases infecting banana production in this region. It is a good idea for using advanced geospatial information obtained from multispectral aerial photographs taken using an unmanned aerial vehicle (UAV), combined with the reliable field data to monitor the incidence of these two diseases. However, the key question is what are the incidences of two diseases in your surveyed field, how reliable of this methodology if you look the Figure 8 which was indicated there was no much healthy plants left. Furthermore, this manuscript is still needed for sufficient revision. Some suggestions and comments indicated below for author’s further improvement.

1. In line 36, change keyword “Blood Diseases Banana” into “Banana Diseases Blood”.

The phrase “Blood Diseases Banana” has already been replaced by “Banana Diseases Blood” in the entire text.

2. Give full name of Fusarium oxysporum f. sp. cubense. Also indicate if infected banana field presented in this manuscript is caused by Race 1 or Tropical Race 4 (TR4).

Fusarium (which is caused by the fungus Fusarium oxysporum forma (f) speciale (sp.) cu-bense is caused by Race 1 or Tropical Race 4 (TR4))

3. In line 42, remove 1996.

The 1996 has been removed.

4. In line 88, correct “Zea may” into “Zea mays“.

      The Zea may have been changed into Zea mays.

5. In line 159-160, give genome composition of four banana cultivars if it is possible.

The genome of four cultivars of bananas has been added. It writes as Pisang Kepok (Musa spp., ABB), Pisang Ambon (Musa acuminata, AAA), Pisang Kapas (Musa spp., AAB), and Pisang Raja (Musa spp., AAB).

6. Add additional Figure to present typical symptom photos of FWB and BDB in individual plant; pseudostem, peduncle and other organs to show what are the main differences between these two main diseases.
7. Give your full description how can you distinguish FWB and BDB.

The following description responds to both questions in numbers 6 and 7

Regarding the Fusarium and BDB-infected banana trees are likely similar, but in detail, both plant diseases can still be distinguished so that the field investigation for data collection can be managed properly. The following visualization can be used as the foundation to differentiate both diseases in Banana trees.

1. Fusarium wilt disease causes pseudostem cracking (yellow arrow in Figure 2b), while a BDB attack does not cause pseudostem cracking (Figure 2a).
2. The symptoms appear on the inflorescence of plants attacked by BDB even though the fruits are still green, but the flowers (male bud) dried (red arrow in Figure 2c). Plants attacked by fusarium wilt generally fail to produce flowers/fruits.
3. BDB attack on fruit causes fruit flesh to rot (Figure 2d), while fusarium wilt attack on mature plants does not cause fruit rot. Fusarium wilt attack on young plants causes plants to die before fruiting.
4. Plants attacked by BDB show symptoms of light-yellow leaves that are evenly distributed and gradually dry out (Figure 2a). While the symptoms of fusarium wilt attack show yellowing starting from the edge of the leaf towards the leaf veins (Figure 2b)

Figure 2 Comparison of individual trees of banana affected by BDB (a, c, and d) and Fusarium (b and d).

8. Linking soil pH to the incidence of Fusarium wilt in lines 358-359 should be placed in the discussion section.
9. This is the same as above to apply for lines 431-435, 440. They should be placed in the discussion section.

All the sentences in lines 358-359, 431-435, and 440 have been moved to the end of the discussion section.

This is similar to what Segura et al. [56] found by taking the Gros Michel bananas (Musa AAA) as samples. Their study found that with an increase in soil pH, a reduction in the incidence of Fusarium wilt occurred in almost all cases. This is why the number of healthy bananas increased and that of affected bananas reduced with a high soil pH. As suggested by Thi et al. [58], Fusarium wilt not only impacts the overall yield during the time of infection but also affects the land used for banana cultivation for the next 20 years. In addition, all affected banana trees may have to be removed to solve the problem [59,60]. However, Fusarium spores will remain in the soil, and as a result, reinfection of new banana accessions in the same area is very likely in the absence of complete soil disinfection [61].

10. Give the full name of MCARI when it is firstly appeared in the manuscript.

the modified chlorophyll absorption in reflectance index as the full name of MCARI has been added into the text.

Author Response File: Author Response.pdf

Reviewer 2 Report

I have carefully assessed the submitted manuscript titled "A Study on the Distribution Pattern of Banana Blood Diseases (BDB) and Fusarium Wilt Using Multispectral Aerial Photo and Handheld Spectrometer in Subang-Indonesia," and I would like to provide a detailed evaluation that I believe will be valuable for the authors and the overall quality of the journal.

Firstly, I would like to draw attention to some crucial linguistic and technical aspects. The terminology "banana blood diseases" appears incorrect, as the singular form "banana blood disease" is more appropriate. Similarly, consistency in acronyms is essential, and "BBD" should be used consistently rather than "BDB."

Furthermore, the manuscript could greatly benefit from an introductory section that adequately introduces the topic of Fusarium wilt of bananas. An absence of this essential context might pose difficulties for readers unfamiliar with the subject matter. Additionally, I recommend addressing the broader context of plant disease distribution and spatial patterns. For instance, the statement in lines 180-182 that a multispectral aerial photograph is mandatory for understanding spatial distribution patterns should be reevaluated. It is important to acknowledge that alternative methods, as exemplified by the works of Madden, Turecheck, Gent, Heck, Pethybridge, and others, have effectively addressed disease spatial patterns without relying solely on multispectral aerial photographs. Moreover, the study by Heck et al. in 2021 demonstrated a successful comprehension of the spatial pattern of Fusarium wilt in bananas without utilizing aerial photographs.

A key concern that emerged from my review pertains to the distinction between "spatial pattern" and "spatial distribution" in the context of plant disease epidemiology. I advise rectifying instances of interchangeably using these terms, such as the reference in line 182. Ensuring conceptual clarity in these matters is essential for precisely understanding the research.

Additionally, the manuscript lacks essential information regarding the validation process of the predictions made. The authors mentioned that NDVI values exhibited discriminatory capabilities among different disease types, yet the number of accurately diagnosed plants is not specified. The confusion matrix in Table 3 indicates the use of only 29 plants for validation purposes, with a breakdown of 13 healthy plants, 5 displaying symptoms of BBD, and 11 with Fusarium wilt of bananas. This is far from ideal.

While I acknowledge the potential significance of the study's concept, I am compelled to express reservations about the clarity of the methods employed and the presentation of equations, which sometimes appear misleading. Notably, the application of machine learning algorithms to predict disease outcomes needs a robust validation dataset. Unfortunately, this fundamental requirement is absent from the manuscript. The absence of solid scientific approaches has led to the propagation of erroneous assumptions within the study.

In conclusion, I appreciate the authors' efforts in addressing the problems caused by BBD and FWB. However, in its current state, the manuscript lacks the requisite clarity in methods, accuracy in equations, and robust validation processes to support its conclusions. I recommend that substantial revisions be undertaken to rectify these issues. Should the authors address these concerns adequately, I believe the manuscript could potentially make a valuable contribution to the field.

Minor English corrections are needed, mainly in the word choices and adjective order. 

Author Response

Comments and Suggestions for Authors

I have carefully assessed the submitted manuscript titled "A Study on the Distribution Pattern of Banana Blood Diseases (BDB) and Fusarium Wilt Using Multispectral Aerial Photo and Handheld Spectrometer in Subang-Indonesia," and I would like to provide a detailed evaluation that I believe will be valuable for the authors and the overall quality of the journal.

Firstly, I would like to draw attention to some crucial linguistic and technical aspects. The terminology "banana blood diseases" appears incorrect, as the singular form "banana blood disease" is more appropriate. Similarly, consistency in acronyms is essential, and "BBD" should be used consistently rather than "BDB."

The phrase "banana blood diseases" have been changed into banana blood disease.

The acronyms of BBD has been used instead BDB

Furthermore, the manuscript could greatly benefit from an introductory section that adequately introduces the topic of Fusarium wilt of bananas. An absence of this essential context might pose difficulties for readers unfamiliar with the subject matter. Additionally, I recommend addressing the broader context of plant disease distribution and spatial patterns. For instance, the statement in lines 180-182 that a multispectral aerial photograph is mandatory for understanding spatial distribution patterns should be reevaluated. It is important to acknowledge that alternative methods, as exemplified by the works of Madden, Turecheck, Gent, Heck, Pethybridge, and others, have effectively addressed disease spatial patterns without relying solely on multispectral aerial photographs. Moreover, the study by Heck et al. in 2021 demonstrated a successful comprehension of the spatial pattern of Fusarium wilt in bananas without utilizing aerial photographs.

An additional description has been added to enhance the explanation of how plant (banana) disease was investigated and factors influencing the disease distribution.

In line 75-77

Some major environmental factors that influence disease distribution have been identified in banana fields in Brazil and are affected predominantly by cultivars, soil physical factors, and field management [6].

In line 85-94

Previously, Fusarium's observation of infected bananas was conducted by Heck et al. [11] to investigate the spatial and temporal dynamics of using georeferenced of observed banana trees in 30 fields. Here, the spatial pattern of infected banana trees is obtained using statistical analysis. Similarly, with Ledesma et al. [12], the improvement was made by accompanying the spatial climatic data. Both studies have successfully shown how Fusarium-infected banana trees can be understood spatially. However, many studies that use spatial data, like satellite imagery from satellite observation, can be considered alternative approaches since it allows to cover a large area and possibly conduct multi-temporal analysis. Additionally, This approach can be combined with various environmental data (e.g. climate, soil, and plant), often using vegetation observation.

A key concern that emerged from my review pertains to the distinction between "spatial pattern" and "spatial distribution" in the context of plant disease epidemiology. I advise rectifying instances of interchangeably using these terms, such as the reference in line 182. Ensuring conceptual clarity in these matters is essential for precisely understanding the research.

Here, we considered replacing the terms "spatial pattern" with "spatial distribution" to achieve consistency, and this study is conducted in a single-day observation instead of performing multi-temporal observation using several aerial photographs.

Additionally, the manuscript lacks essential information regarding the validation process of the predictions made. The authors mentioned that NDVI values exhibited discriminatory capabilities among different disease types, yet the number of accurately diagnosed plants is not specified. The confusion matrix in Table 3 indicates the use of only 29 plants for validation purposes, with a breakdown of 13 healthy plants, 5 displaying symptoms of BBD, and 11 with Fusarium wilt of bananas. This is far from ideal.

A perfect classification accuracy at 100% achieved using an RF algorithm is rare. Studies are using large training sample sizes, while at the same time considering their critical and optimal size [60]. Of course, the critical must be lowest compared to the opti-mal. Since this study only depends on the 29 samples, it was certainly limited. However, in some cases explained by Luan et al. [61], the predictive performance of the RF model was substantially improved when the sample size increased from 10 to 30 sites, but less improvement was evident with larger datasets. We ensure that when the sample size used for running the RF is decreased to the smallest size (e.g., 5, 10, 15, and 20), and this sample size’s class attribute corresponds to all plant diseases, the classification accuracy might be lower compared to maximum collected samples size [62]. It will not continuously im-prove the accuracy once it uses the largest sample size. On the other hand, the increasing sample size can improve the predictive power of the RF models, but the effects are con-strained. On the other hand, implementing the RF algorithm is still possible. Although the sample size is limited, the RF model may benefit from improving predictive performance from an ensemble of basic tree learners from small datasets [63].

While I acknowledge the potential significance of the study's concept, I am compelled to express reservations about the clarity of the methods employed and the presentation of equations, which sometimes appear misleading. Notably, the application of machine learning algorithms to predict disease outcomes needs a robust validation dataset. Unfortunately, this fundamental requirement is absent from the manuscript. The absence of solid scientific approaches has led to the propagation of erroneous assumptions within the study.

In conclusion, I appreciate the authors' efforts in addressing the problems caused by BBD and FWB. However, in its current state, the manuscript lacks the requisite clarity in methods, accuracy in equations, and robust validation processes to support its conclusions. I recommend that substantial revisions be undertaken to rectify these issues. Should the authors address these concerns adequately, I believe the manuscript could potentially make a valuable contribution to the field.

Author Response File: Author Response.pdf