Drivers of Long-Term Land-Use Pressure in the Merguellil Wadi, Tunisia, Using DPSIR Approach and Remote Sensing

Round 1

Reviewer 1 Report

The manuscript has been improved according to the observation. Good job! Congratulations.   

Author Response

We would like to express our sincere thanks to the editor and the reviewers for the valuable comments and time spent to improve the manuscript. We now firmly believe that with the implemented changes in text, the manuscript has improved greatly and would be of great interest to the readers of the Land journal. We have modified according to all comments made by the reviewers. All modifications are highlighted in green in the manuscript. Please find below the reviewer´s comments and our answer and changes in the green text below each comment.

The manuscript has been improved according to the observation. Good job! Congratulations.   

The authors thank the reviewer for positive remarks.

Reviewer 2 Report

The authors have responded all review comments properly

Author Response

We would like to express our sincere thanks to the editor and the reviewers for the valuable comments and time spent to improve the manuscript. We now firmly believe that with the implemented changes in text, the manuscript has improved greatly and would be of great interest to the readers of the Land journal. We have modified according to all comments made by the reviewers. All modifications are highlighted in green in the manuscript. Please find below the reviewer´s comments and our answer and changes in the green text below each comment.

The authors have responded all review comments properly.

The authors thank the reviewer for positive remarks.

Reviewer 3 Report

Although the document was well presented, there are some problems in the methodology related to the mapping of land use and land cover and in the direct correlation of its dynamics with the determinants presented.
The individual mapping protocol of the LULC maps generated significant variations in the global accuracy or in the thematic classes of each one of them, making it difficult or even preventing their comparison over time, without any consideration related to this heterogeneity.
The identification, characterization and correlations of drivers, pressures, conditions, impacts and responses to adverse changes in land cover seems to have been performed in a very intuitive and unscientific way.
There are important adjustments to the adequacy of the land use and land cover mapping methodology, with the aim of obtaining more consistent and homogeneous results throughout the series.
Furthermore, the protocol for obtaining correlations between LULC dynamics, specifically between 1976 and 2016, and its complex interactions with social, economic and biophysical data needs to be statistically explained.

Author Response

We would like to express our sincere thanks to the editor and the reviewers for the valuable comments and time spent to improve the manuscript. We now firmly believe that with the implemented changes in text, the manuscript has improved greatly and would be of great interest to the readers of the Land journal. We have modified according to all comments made by the reviewers. All modifications are highlighted in green in the manuscript. Please find below the reviewer´s comments and our answer and changes in the green text below each comment.

1. Although the document was well presented, there are some problems in the methodology related to the mapping of land use and land cover and in the direct correlation of its dynamics with the determinants presented.

We have now done our best to improve the description of methodology. Thus, we have added new text in section 2.6; lines 233 – 254 and section 3.3; lines 364 – 374, this includes the correlation analysis (R) to determine relationships between land use / land cover dynamics and the DPSIR approach. Figure 4 illustrates the relationship between the probability of occurrence for the driving forces of arboriculture dynamics as perceived by households. We denote by (1) Population growth; (2) Creation of douars; (3) Climate change; (4) Creation of agricultural plots; (5) Land fragmentation; and (6) Mountainous and rugged topography.

Figure 4. Relationship between probability of occurrence for driving forces of arboriculture dynamics as perceived by the households.

Determination of the relationship between land use land cover types of culture and the DPSIR approach

Table 11 illustrates the strength of the relationships between each land use land cover type and the probability of occurrence for the DPSIR approach of LULC types of culture dynamics as perceived by households.

Table 11. Determination coefficient (R²) for LULC types of culture and the probability of occurrence for the DPSIR approach of LULC types of culture dynamics as perceived by households.

The determination coefficient is on average equal to 0.76. This shows that there is a strong correlation between the land cover class and the different components of the DPSIR approach.

2. The individual mapping protocol of the LULC maps generated significant variations in the global accuracy or in the thematic classes of each one of them, making it difficult or even preventing their comparison over time, without any consideration related to this heterogeneity.

We understand the question and it is very relevant. The rate of the overall accuracy is related to the resolution of the satellite images used that in this case was 30 m using the Landsat images. The same type of images has been used in numerous other published studies with similar results in terms of accuracy (e.g., Gedefaw et al., 2020) (Section 3.1 lines 262,263). The results of these images are consistent over time. As the correlation analyses above shows (Table 2), coefficients of determination are generally very high and statistically significant. Thus, not only individual maps but as well the collective group of maps show statistically significant relationships in terms of the investigated variables.  

3. 
   The identification, characterization and correlations of drivers, pressures, conditions, impacts and responses to adverse changes in land cover seems to have been performed in a very intuitive and unscientific way.

We thank the reviewer for time spent to improve the quality of the manuscript, however, the authors have difficulties in perceiving this statement. The study follows a pre-defined scientific methodology that is described in detail and that results in statistically measurable results and cause - effect relationships that arise out of interactions among physical, economic, and social components. The DPSIR approach is clearly not unscientific. The study is inspired by other scientific studies that have related remote sensing with the indicators of the DPSIR approach (e.g., Gedefaw et al., 2020; Rasool et al., 2021). These and other studies are not intuitive and unscientific. In a broader scientific sense, statistical inference is used.

4. 
   There are important adjustments to the adequacy of the land use and land cover mapping methodology, with the aim of obtaining more consistent and homogeneous results throughout the series.

The reviewer states adjustments, however, does clarify what adjustments are deemed necessary. Thus, it is difficult to reply to this criticism. In our view, the best “adjustment” is to perform field studies to quantify the accuracy of satellite imagery. This, we have performed. Besides, the surface characteristics of the Merguellil Wadi Basin is one of most studied in Tunisia. The basin surface characteristics have been monitored through a national inventory from 1995 to 2010.  

5. 
   Furthermore, the protocol for obtaining correlations between LULC dynamics, specifically between 1976 and 2016, and its complex interactions with social, economic and biophysical data needs to be statistically explained.

We understand the concern of the reviewer and for this purpose we have added the correlation analyses as outlined above. The statistical results using inference show that the results are robust and not just by coincidence.  Section 3.3, lines 364 – 374.

Author Response File: Author Response.docx

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Some comments about the manuscript:

Line 14: The study presents… (change sentence) 

Line 15: The main objective was… (It is part of the methodology, please improve the sentences)

Line 24: You say that rural residents are guilty of damage. What about of “urban residents”? What should they do to solve the problems? The solution would be integral.

Line 39: Change various by several (or synonyms)

Lines 92-94: Please improve your writing skills. Review all the manuscript.

Line 99: Maybe change “Experimental areas” to “study area” (it is most conventional)

Lines 117-118. It seems that you have used an English translator directly. Please, improve the text wording.

Lines 238-243: These lines describe Table 4. A table must be understood by itself. You do not need to describe.

Some comments about the content:

The classification of land-use change and its temporal analysis is adequately performed. 

Drivers need more evidence and analysis. They have a subjective analysis. More information needs to be gathered.

Similarly, the effects on society and the landscape are not strong and do not have a solid base to obtain robust results. More information and analysis are needed. 

English translation needs to be improved.

Reviewer 2 Report

This research studies 40 years land use change in Tunisia using remote sensing data and  investigates the possible drivers behind the increased pressure. The key findings of the study are 1), during the last 40 years, arboriculture, cultivated land and bare soil areas increased and, 2) the economic development, cultivation practices and hydroagricultual techniques are the important drives for the increase.  

The paper is professionally written and structured. The study fits well into the scope of the journal and makes a significant contribution to the community. The manuscript will be suitable for publication once a few issues as discussed in the following have been considered. 

SPECIFIC COMMENT 

1. The authors use three single images to conduct the land useclassification. Thisapproach could be problematic if the selected images are not suitable for such study, like cloud/haze/smoke influence, as well as some short-lived features (burned area, moisture changes, flood...).  You can see that the overall accuracy for classification is not very high (~0.8). Moreover, the classification of different types of vegetation  are very complicated, it would be helpful if the authors could provide confusion matrix.  
2. In the second part of the study, the authors state that the key drivers areeconomicdevelopment, cultivation practices and hydro agricultural techniques, but very few statistics information was provided. The link between these factors and the changes are not built. 

Reviewer 3 Report

Reviewer’s comments

Title: Drivers of Long-term Land-Use Pressure in the Merguellil Wa-2 di, Tunisia, using DPSIR Approach and Remote Sensing

Major comments

1. In Chapter 2, it is considered that data period (20-year) between specific times to examine the land use dynamics. Is there a standard set at 20-year intervals?

2. In Chapter 2, the process of accuracy assessment is important step. So, descript the accuracy assessment in details in this chapter.

3. In Chapter 2, is it possible that if only the latest data (2016’s) are reflected in the household surveys, not old data? And is it possible to quantitatively analyze the effect of each step on each other in the D-S-P-I-R approach?

4. In Chapter 3, because the period (20-year) between data analyzed in land use is long, it is difficult to explain that the analysis results of land use show a specific trend. It is considered that more land use data are needed for the land use analysis.

5. In Chapter 3, this paper explained the relationships between two factors or three factors (land use change – climate (change) impact (risk) – industry (agriculture) change) using the DSPIR approach. However, it is required for the more detailed explanation of relationship of three factors.

6. It is recommended that results of Chapter 4.2 are presented briefly in a table.