Compatible Biomass Model of Moso Bamboo with Measurement Error
, Fengying Guan 1,2,*, Xiao Xiao 1,2, Xuan Zhang 1,2 and Chengji Li 1,2Round 1
Reviewer 1 Report
Journal: Forests (ISSN 1999-4907)
Manuscript ID: forests- 1719394
Title: Compatible biomass model of Moso bamboo with measurement error
Overall Comments and Suggestions for Authors
Dear author,
Regarding the biomass model of bamboo for prediction especially in Jiangsu region of China, this manuscript may be interesting to the relevant researchers who deals with similar issues such as forest biometrician, growth and yield modeler, and bamboo ecologist and silviculturist. I consider the overall structure and flow are fine. Especially, Introduction and Materials Methods were clearly explained. However, some results and explanation can be required for better description of this manuscript. From my point of view, I made several comments to improve the manuscript as below. I wish the developed models in this study will serve for managing bamboo well in the region.
I hope that this manuscript can be improved based on peer-review’s comments. My specific comments were provided in detail as follows.
Kind regards,
Reviewer
Point 1.
In Materials and Method, is the stem form of bamboo not affected by stand condition? If stem forms are different by the bamboo part depending on the stand condition such as stand density, more description of stand condition should be added in the Materials section. It will be helpful for the model application.
Point 2.
I am a bit wondering about the sample size of this study. Regarding the target species, do authors consider that the number of tree samples are sufficient for unbiased prediction in the field? It would be better to cite some similar previous studies regarding this issue if possible. It can help to support the ideas and the sample size of this study.
Point 3.
I consider that the current format of the description in the Results section is not readable to the readers. I don’t think it’s needed to write the same equation all the time. What about mentioning the equations directly with parameter estimates? For example, once it’s introduced in Methods, the final equations can be presented with the parameter input of Table 2.
Point 4.
In the Results section, I think authors need to provide the significance of each parameter estimates with t-value and/or p-value. If not preferred because of the manuscript style, the clear explanation must be included in the text of the Results section.
Point 5.
In the Results section, can authors add a chapter to introduce how to apply the model finally suggested by authors? It will help stakeholders to apply.
Point 6.
In the Discussion section, the comparison between modelling approaches was discussed well. In addition to this, I recommend authors to write the pros and cons or restrictions for practical use. For example, authors can add some cautions of spatial and/or temporal range for application. It’s because some input data range can be extrapolated and thus make some bias. Also, it can be about the range or characteristics of input variables such as D, H, and NLDBH.
Minor comments.
Line 8: What is this affiliation? I cannot see this from the author list.
Line 12: Tree-level sample? It should be clarified. Mention “tree” clearly with the scientific name.
Line 15: what are the “these” six variables? Remove or revise the phrase.
Table 2. What is “A”? Age? define it clearly.
Correct the decimal point and unify it by variables. It would be more significant. For example, Was dbh significantly measured with three decimal points?
Author Response
Reviewer 1
Comments and Suggestions for Authors
Journal: Forests (ISSN 1999-4907)
Manuscript ID: forests- 1719394
Title: Compatible biomass model of Moso bamboo with measurement error
Overall Comments and Suggestions for Authors
Dear author,
Regarding the biomass model of bamboo for prediction especially in Jiangsu region of China, this manuscript may be interesting to the relevant researchers who deals with similar issues such as forest biometrician, growth and yield modeler, and bamboo ecologist and silviculturist. I consider the overall structure and flow are fine. Especially, Introduction and Materials Methods were clearly explained. However, some results and explanation can be required for better description of this manuscript. From my point of view, I made several comments to improve the manuscript as below. I wish the developed models in this study will serve for managing bamboo well in the region.
I hope that this manuscript can be improved based on peer-review’s comments. My specific comments were provided in detail as follows.
Kind regards,
Reviewer
Point 1.
In Materials and Method, is the stem form of bamboo not affected by stand condition? If stem forms are different by the bamboo part depending on the stand condition such as stand density, more description of stand condition should be added in the Materials section. It will be helpful for the model application.
Response: thanks for your advice. We added the stand condition in the manuscript. See Line 90-97.
The main types of traditional management measures adopted in Moso bamboo forests in this region include bamboo forest cutting (cutting bamboo over three degrees old), shrub and grass cutting, winter bamboo shoot and spring bamboo shoot harvesting, bamboo cutting, and tourism. Yixing forest farms do not fertilize Moso bamboo forests. All human behavior in the experimental area was controlled. The stand density of Moso bamboo was 2000–4000 plants/hm2, the average DBH was approximately 9.8 cm, and the age structure of grade I, grade II, and grade III bamboo was 3:4:3.
Point 2.
I am a bit wondering about the sample size of this study. Regarding the target species, do authors consider that the number of tree samples are sufficient for unbiased prediction in the field? It would be better to cite some similar previous studies regarding this issue if possible. It can help to support the ideas and the sample size of this study.
Response: thanks for your advice. We added some similar previous studies in the manuscript. See Line 200-204.
Some studies have proposed that the parameter estimation is unbiased when the sample size is 50 [37-39]. In the study of bamboo forest biomass, Yang (2016) and Guo et al. (2015) established biomass models using the measured data of 32 and 25 moso bamboos, respectively [40,41]. Therefore, the sample data size estimation in this study was unbiased.
Point 3.
I consider that the current format of the description in the Results section is not readable to the readers. I don’t think it’s needed to write the same equation all the time. What about mentioning the equations directly with parameter estimates? For example, once it’s introduced in Methods, the final equations can be presented with the parameter input of Table 2.
Response: thanks for your advice. We modified this in the manuscript. Please see result in the manuscript. See line 226-344.
Point 4.
In the Results section, I think authors need to provide the significance of each parameter estimates with t-value and/or p-value. If not preferred because of the manuscript style, the clear explanation must be included in the text of the Results section.
Response: thanks for your advice. We added this in the manuscript.
Since there is no t value or P value calculated by Forstate, CV is used in this study. See Line 195-200.
Please see result in the manuscript.
See line 196-200. See table 3, 5 and 7.
In addition, the fitted model also requires stable parameters (t value of the estimated value of the parameter > 2 or the coefficient of variation (CV) < 50%) and a random residual distribution (the positive and negative residuals of each diameter order were offset, and 0 was used as the reference line and distributed symmetrically below). Therefore, the stability of parameter estimation in this study mainly refers to the CV.
Point 5.
In the Results section, can authors add a chapter to introduce how to apply the model finally suggested by authors? It will help stakeholders to apply.
Response: thanks for your advice. We added this in the manuscript.
See line 339-344.
The one-, two-, and three-prediction variable model systems estimated by the NEIVM one-step method proposed in this study can calculate the aboveground biomass and that of each component. Under the same site conditions as the study area, the stand density was 2000–3500 and the average DBH was 8–10 cm. The biomass value re-quired by the operator can be obtained from the collected bamboo forest factors com-bined with the model form in this study.
Point 6.
In the Discussion section, the comparison between modelling approaches was discussed well. In addition to this, I recommend authors to write the pros and cons or restrictions for practical use. For example, authors can add some cautions of spatial and/or temporal range for application. It’s because some input data range can be extrapolated and thus make some bias. Also, it can be about the range or characteristics of input variables such as D, H, and NLDBH.
Response: thanks for your advice. We added this in the manuscript.
See line 395-409.
The TSEM one-step, two-step, and NSUR proportional methods calculate AG biomass first, and then distribute it to each component; however, the summation method of NSUR calculates each component first, and then sums it. In forestry practice, the goal is to obtain the whole plant or aboveground biomass [47,48]. Therefore, it is more practical to choose a biomass model estimated using the TSEM one-step method.
Owing to the operational needs of this area, there was no bamboo older than 4 degrees in the bamboo forest. If bamboo older than 4 degrees is included in the calculation, there may be a certain deviation in the calculation of the biomass. When the site condi-tions to be calculated are similar to those of this study (stand density of 2000–3500 and the average DBH of 8–10 cm), the model can be used directly for calculation.
However, the application of biomass models has limitations in terms of scale or region. Moso bamboo grown in different regions often exhibits differences in stem shape and biomass distribution. When the biomass model established on a small scale or in a small watershed is extrapolated to a larger scale or regions, samples need to be collected for model verification and parameter adjustment.
Minor comments.
Line 8: What is this affiliation? I cannot see this from the author list.
Response: thanks for your advice. We deleted the line in the manuscript.
Line 12: Tree-level sample? It should be clarified. Mention “tree” clearly with the scientific name.
Response: thanks for your advice. We modified this in the manuscript.
See Line 13-16.
Based on the measured individual biomass data of 66 Phyllostachys heterocycla cv. Pubescens plants in the Yixing state-owned forest in Jiangsu Province, nonlinear simultaneous equations with meas-urement errors were constructed using nonlinear error-in-variable models (NEIVM) (one step, two step) and non-linear seemingly unrelated regression (NSUR).
Line 15: what are the “these” six variables? Remove or revise the phrase.
Response: thanks for your advice. We modified this in the manuscript.
See Line 16-19.
Variables affecting biomass were evaluated, including diameter at breast height (DBH), bamboo height (H), height to crown base (HCB), node length at DBH (NL), base diameter (BD), and bamboo age (A). DBH, H, and HCB had significant effects on the biomass of each component.
Table 2. What is “A”? Age? define it clearly.
Response: thanks for your advice. We added this in the manuscript.
See Line 111-113.
Bamboo age (A) was determined based on the features of the culms, such as external color, branch, and leaf development, and the status of the culm sheaths.
Correct the decimal point and unify it by variables. It would be more significant. For example, Was dbh significantly measured with three decimal points?
Response: thanks for your advice. We modified this in the manuscript.
DBH is measured in two decimal places. In order to ensure uniformity, four decimal places are reserved in this study.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Line 8: No. 3 is missing from the authors' list. Delete the line or add the number to the authors' list.
Line 15: “The effects of these six variables on biomass were also evaluated.” Which variables? You did not provide information on the six variables, so I do not know what you evaluated. Please, recast that sentence.
Line 85: rearrange the coordinates. North (N) should come before East (E).
Line 99-101: What sampling method did you use to select the 66 bamboo samples? You wrote, “Using the typical sample selection method”. To my knowledge, this is not a sampling method.
Line 104-105: How did you sample the branches and leaves? Please, provide a detailed sampling procedure.
Line 119-121 (Table 1): What is the difference between diameter at chest height (D) and DBH? Conventionally, DBH is the diameter at breast height. I suggest you do some changes to the abbreviations. Maybe, D: diameter at breast height, NL: node length at breast height. What does the abbreviation A stand for? You need to define it.
Line 199: Provide the R code as an appendix
Line 113-114 (Fig 2): Include the correlation and the p-values to the plots
Line 201-205: The authors wrote “… we found that D, H, and A had significant effects on the total aboveground biomass and component biomass. Therefore, by using these three variables, we developed a one-parameter predictor system with only D, a two-predictor system with D and H, and a three-predictor system with D, H, and NLDBH.” I am a bit surprised that variable A (which was never defined in the method section) was not used. Yet the authors said “these three variables”. Excuse me, why would you include NLDBH when its contribution was not significant (based on your first statement). A closer look at Fig 2 (Line 113-144) showed that NLDBH had no significant relationship with total biomass (AG), Stem biomass, Branch biomass, and Leaf biomass. Include the scatterplot of variable A with the total AG and biomass components in Fig 2. Please, revisit your models and provide clarification.
Line 218-219: The indices are relatively the same up to 3 decimal places. For me, NEIVM is not handy and provides little or no improvement.
Line 298: Change the font colour.
Line 322: Replace “… is discovered that DBH had the greatest impact …” with ‘is discovered that diameter at breast height (D) had the greatest impact’.
Author Response
Reviewer 2
Comments and Suggestions for Authors
Line 8: No. 3 is missing from the authors' list. Delete the line or add the number to the authors' list.
Response: thanks for your advice. We deleted the line in the manuscript.
Line 15: “The effects of these six variables on biomass were also evaluated.” Which variables? You did not provide information on the six variables, so I do not know what you evaluated. Please, recast that sentence.
Response: thanks for your advice. We modified this sentence.
See Line 16-19.
Variables affecting biomass were evaluated, including diameter at breast height (DBH), bamboo height (H), height to crown base (HCB), node length at DBH (NL), base diameter (BD), and bamboo age (A). DBH, H, and HCB had significant effects on the biomass of each component.
Line 85: rearrange the coordinates. North (N) should come before East (E).
Response: thanks for your advice. We modified this. See Line 85-86.
The study area is located on the Yixing state-owned forest farm, Jiangsu Province, western China (31°13'–31°15' N, 119°41'–119°44'E).
Line 99-101: What sampling method did you use to select the 66 bamboo samples? You wrote, “Using the typical sample selection method”. To my knowledge, this is not a sampling method.
Response: thanks for your advice. We modified this.
See Line 103-109.
First, bamboo in the study area was measured to obtain the DBH distribution of the bamboo forest in the study area. Then, according to the two factors of DBH (7–14 cm) and age (1–4 degrees), a stratified sampling method was adopted. Taking 1 cm as the diameter class, 10 stems were sampled from each diameter class. Because there was lit-tle bamboo in the large-diameter and small-diameter classes, eight stems were taken for each diameter class. A total of 66 bamboo samples were obtained.
Line 104-105: How did you sample the branches and leaves? Please, provide a detailed sampling procedure.
Response: thanks for your advice. We modified this part in the manuscript. See Line 113-124.
The bamboo culm biomass was sampled, and the fresh weight of each section of bamboo was determined using the full weight method, marked, and recorded. Samples were taken from the upper, middle, and lower parts of each 2 m section of the bamboo trunk, and 500–1000 g was sampled after mixing. The standard branch method was adopted for the determination of bamboo branch biomass. After the bamboo stem was divided into 2 m sections, three standard branches were selected according to the average base diameter and length in each section, the fresh weight was weighed, the average value was obtained, and the number of live branches in the section was multi-plied to obtain the total fresh weight of the branches in the section. After mixing the standard branches, samples of 500–1000 g were obtained. For the biomass of bamboo leaves, all the leaves of the selected standard branches were removed, the fresh weight was obtained, and 100–200 g was taken as a sample and taken back to the laboratory.
Line 119-121 (Table 1): What is the difference between diameter at chest height (D) and DBH? Conventionally, DBH is the diameter at breast height. I suggest you do some changes to the abbreviations. Maybe, D: diameter at breast height, NL: node length at breast height. What does the abbreviation A stand for? You need to define it.
Response: thanks for your advice. We modified this part in the manuscript. Diameter at chest height (D) and DBH were no difference, and we modified this (Unified use of DBH, and replace chest with breast). See Line 138-140.
(notes: diameter at breast height (DBH), bamboo height (H), height to crown base (HCB), bamboo age (A), node length at DBH (NL), base diameter (BD), and total aboveground (AG) biomass)
Line 199: Provide the R code as an appendix
Response: thanks for your advice. We Provide the R code as an appendix.
Line 113-114 (Fig 2): Include the correlation and the p-values to the plots.
Response: thanks for your advice. We added this in the Fig2.
Since the p value is small, the 4 decimal places may be 0, so only the correlation coefficient is shown in the figure. The p value is also explained in the manuscript.
Line 201-205: The authors wrote “… we found that D, H, and A had significant effects on the total aboveground biomass and component biomass. Therefore, by using these three variables, we developed a one-parameter predictor system with only D, a two-predictor system with D and H, and a three-predictor system with D, H, and NLDBH.” I am a bit surprised that variable A (which was never defined in the method section) was not used. Yet the authors said “these three variables”. Excuse me, why would you include NLDBH when its contribution was not significant (based on your first statement). A closer look at Fig 2 (Line 113-144) showed that NLDBH had no significant relationship with total biomass (AG), Stem biomass, Branch biomass, and Leaf biomass. Include the scatterplot of variable A with the total AG and biomass components in Fig 2. Please, revisit your models and provide clarification.
Response: thanks for your advice. We modified this in the manuscript.
This sentence is my mistake in writing. It should be this that by comparing the six independent variables, we found that D, H, and HCB had significant effects on total aboveground biomass and component biomass (Fig 2.). The strongest correlation in the study should be DBH, H and BD. However, because DBH and BD have significant collinearity, HCB is selected for replacement.
See Line 227-232.
By comparing the six independent variables, we found that DBH, H, BD, and HCB had significant effects on the total aboveground biomass and component biomass (Fig 2). Because DBH and BD have strong collinearity, only one was considered in this study. Therefore, using three variables (DBH, H, and HCB), we developed a one-predictor system with only DBH, a two-predictor system with DBH and H, and a three-predictor system with DBH, H, and HCB. Table 2 shows the parameters of the basic model.
Line 218-219: The indices are relatively the same up to 3 decimal places. For me, NEIVM is not handy and provides little or no improvement.
Response：thanks for your advice.
This study only provides some comparison of methods, only considering the measurement error and the difference of results caused by different parameter estimation methods. In some previous studies, different parameter estimation methods have little impact on the results for the same model form. For example Fu et al 2016, Liu et al., 2020.
Line 298: Change the font colour.
Response：thanks for your advice. We modified this in the manuscript. See Line 359.
Line 322: Replace “… is discovered that DBH had the greatest impact …” with ‘is discovered that diameter at breast height (D) had the greatest impact’.
Response：thanks for your advice. We modified this in the manuscript. See Line 382-383.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Journal: Forests (ISSN 1999-4907)
Manuscript ID: forests- 1719394
Title: Compatible biomass model of Moso bamboo with measurement error
Overall Comments and Suggestions for Authors
Dear authors,
Thank you for your response. I checked all the revisions and I considered that it has been updated much based on the previous comments. I think it can be proceeded for publication.
I appreciate author’s effort on this research.
Kind regards,
Reviewer
