Round 1

Reviewer 1 Report

This work used the data derived from Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and Landsat 8 Operational Land Imager (OLI) scenes to determine vegetation index (VIs) and collected field data of total understorey (live and dead, all growth forms), live understorey (all growth forms), and grass biomass semiarid woodlands of south-eastern Australia, with aims to develop prediction model of biomass using VIs as the predictor variables. Thhe results showed that brightness was an important variable in the semi-arid environment, but in order to improve the explanatory power of models and to predict biomass adequately in new sites, other variables (such plant cover and soil moisture) should be included in the models. The findings could provide valuable information for sustainable management of semi-arid vegetation and vertebrate herbivore population conservation. Overall, this manuscript is well written and the data are detail for model development and validation. However, there are several following major concerns for improving the manuscript in the current form.

1) In the Introduction section, what the achievements in the literature are not deeply reviewed and the gaps needed to be addressed in this study are not clearly presented. Thus, this study suffers from the lack of novelty.

2) The objectives are not good enough if only to develop three types of models. It would be better if the authors revise them as (1) to determine which VIs should be included for developing the models to predict understory plant biomass, and (2) to examine whether the models could be improved and extended to other site when including plant and soil variables.

3) In the Discussion section, the question of which variables and why are included in the model is not logically discussed. Please follow a pipeline from remote sensing data, to plant data, and soil data, to discuss they are best fitted or not, why.

4) Other specific comments please see the remarks on the attached pdf file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Overall a very interesting paper. The results do not surprise me. But having said that, this is nevertheless useful in the sense that it provides evidence of the limitations of RS and ratios. The paper is very well written, structured and presented and is very informative. The application I found particularly interesting. The methodology etc. seems sound and the summary and conclusions sensible. Very minor things:

1) Perhaps an additional read over to pickup a few minor text edits: e.g. sentence such as this... As this threshold is based on grass biomass, the to reliably estimate grass biomass is particularly important in this context.

2) Maybe add a key/legend for the mosaic of vegetation in the satellite image (which satellite, date etc.)?

3) Maybe add a Workflow diagram for the methodology as a graphic summary.

4) Some ground based photos of the vegetation types, and the vertical canopy/structure etc. would be informative.

A pleasure to read though. Lidar might be something to consider - although not a practical. Maybe some drone-related LIdar would provide additional insight.

Author Response

Response to Reviewer 2 Comments

Comments and suggestions for authors:

General comment 1: Overall a very interesting paper. The results do not surprise me. But having said that, this is nevertheless useful in the sense that it provides evidence of the limitations of RS and ratios. The paper is very well written, structured and presented and is very informative. The application I found particularly interesting. The methodology etc. seems sound and the summary and conclusions sensible. Very minor things:

We thank the reviewer for taking the time to review our manuscript and provide constructive feedback.

Point 1: Perhaps an additional read over to pickup a few minor text edits: e.g. sentence such as this... As this threshold is based on grass biomass, the to reliably estimate grass biomass is particularly important in this context.

We have re-read the manuscript and corrected any spelling and grammatical errors. In the particular example above, we have inserted the word “ability” between the words “the to”.

Point 2: Maybe add a key/legend for the mosaic of vegetation in the satellite image (which satellite, date etc.)?

This map (Figure 1) was made in QGIS, and the background image is a Google Satellite XYZ tile basemap. No date was available for the basemap in QGIS. We have also included images of the vegetation types sampled demonstrate the variation in vegetation structure and cover across the study area, as suggested in Point 4 below. The photos were all taken during the in December 2017 field campaign.

Consequently, we have reworded the caption: “Figure 1. (a) Map of the Pine Plains area of Wyperfeld National Park, south-eastern Australia (inset). The background image is a Google Satellite XYZ tile basemap. (b) Images captured in December (summer) 2017, which highlight the variation in structure and composition of the vegetation communities sampled.”.

Point 3: Maybe add a Workflow diagram for the methodology as a graphic summary.

We have included an additional figure (Figure 2) which visually outlines our methodology.

Point 4: Some ground based photos of the vegetation types, and the vertical canopy/structure etc. would be informative.

We have included images of four sites, representing each vegetation type sampled, in Figure 1 (refer to Point 2).

General comment 2: A pleasure to read though. Lidar might be something to consider - although not a practical. Maybe some drone-related LIdar would provide additional insight.

Thank you. Although we alluded to the potential of LIDAR in the final paragraph of our discussion, we did not specify drone-related data. We agree this would be useful, and have added the following: “..., such as those obtained using unmanned aerial vehicles (UAVs).”.

Reviewer 3 Report

Dear Authors,

Congrats! You did a great job!

The only point is that please write about research limitations in the conclusion.

Best regards,

Author Response

We thank the reviewer for taking the time to review our manuscript. We have highlighted some key limitations in the discussion and conclusions sections of the paper, namely the challenge of remote sensing model transferability, the limited timeframe of the study, the spatial and temporal resolution of satellite imagery, and the lack of herbivory data.

Round 2

Reviewer 1 Report

The authors have addressed most concerns and revised the manuscript accordingly. I recommend to accept the manuscript.