Evaluating the Effect of Prosopis juliflora, an Alien Invasive Species, on Land Cover Change Using Remote Sensing Approach

Round 1

Reviewer 1 Report

The paper presents a classification of Landsat images to obtain land use categories, which are related to ground data about presence of an invasive species. Classification of Landsat TM images using maximum likelihood algorithm and land use change detection calculated as difference of images do not imply any methodological innovation. Since the proposed method do not provide any innovation or new knowledge, the study results should offer new knowledge about the problem of IPS invasion. However, most of the conclusions and of the result/discussion correspond to other studies and are not drawn from the results of this study. If new knowledge is provided, the paper should explain and highlight it clearer.

General comments:

• The use of Landsat images is justified in section 3.1. However, there is no reference to Sentinel images, which have more spatial resolution, similar radiometric resolution and are open access. Why are Landsat images preferred to Sentinel images?
• How do you explain the decrease of urban area? It is a strange phenomenon which should be justified.
• According to ground data, the 70% of vegetation is IPS, however, the combination of ground data with the results of image classification is very limited; it is reduced to this relationship. In addition to this data (70%), more integration of both data sources could be developed. For example, it could be tested if there is the same proportion of IPS in new vegetation areas and in areas with vegetation since 1999 of if there is a relationship between the increase of vegetation area and the increase of IPS invasion,…. The methodological innovation of this paper could be in the combination of both data sources, however the current development of this integration is too limited.
• Figure 5. If 70% is the data for 2017, how was de % of invaded areas in 1999-2013 calculated?
• In section 3.2:
  • Lines 239-243 and 262-271 are not results, they should be in the introduction
  • Lines 274-313 are not results of this study but of other studies.
• What are the data of Table 4 used for? Which conclusions are obtained from them? Could these data be combined with image classifications results, for example, to estimate IPS biomass?

Author Response

Thank you for the constructive and thoughtful comments. We have revised the manuscript comments and suggestions. The revised text in the new version of the manuscript is in red color. As outlined below, we addressed all your concern points.  We feel the revisions incorporated in the manuscript have strengthened the manuscript and hope you will now find it suitable for publication.

The paper presents a classification of Landsat images to obtain land use categories, which are related to ground data about presence of an invasive species. Classification of Landsat TM images using maximum likelihood algorithm and land use change detection calculated as difference of images do not imply any methodological innovation. Since the proposed method do not provide any innovation or new knowledge, the study results should offer new knowledge about the problem of IPS invasion. However, most of the conclusions and of the result/discussion correspond to other studies and are not drawn from the results of this study. If new knowledge is provided, the paper should explain and highlight it clearer.

Our response: We agree. The results and discussion were separated in the revised version and we focus in-depth on the results of our study.   

General comments:

The use of Landsat images is justified in section 3.1. However, there is no reference to Sentinel images, which have more spatial resolution, similar radiometric resolution and are open access. Why are Landsat images preferred to Sentinel images?

Our response: Sentinel-1 and Sentinel-2A were launched in 2014-15. Our study need data for 1999-2017 which not available from Sentinel sensors.

How do you explain the decrease of urban area? It is a strange phenomenon which should be justified.

Our response: We agree. We added the following paragraph to the discussion section (revised version) to clarify this point. “

The primary source of income for Sweimeh village is the agriculture sector. However, the significant increase in Jordan population (4.4 to 9.8 million) between 2000 and 2019 increased the percentage of people living in water scarce areas by 64% [30]. High demand for water coupled with a reduction in governmental subsidies for agriculture in the last two decades lead farmers to abandon their lands. In Sweimeh the total population decreased by 40% between 2009 and 2015 (6500 vs 5000) [31]. That populations decreased from 2009 to 2015 is consistent with our results which showed that urban area decreased by 11% from 1999 to 2017 (Figures 3 and 4)”

According to ground data, the 70% of vegetation is IPS, however, the combination of ground data with the results of image classification is very limited; it is reduced to this relationship. In addition to this data (70%), more integration of both data sources could be developed. For example, it could be tested if there is the same proportion of IPS in new vegetation areas and in areas with vegetation since 1999 of if there is a relationship between the increase of vegetation area and the increase of IPS invasion,…. The methodological innovation of this paper could be in the combination of both data sources, however the current development of this integration is too limited.

Our response: Interesting point. We revised Figure 2 following the reviewer suggestion. We collected ground reference points using the historical data from the Royal Society for the Conservation of Nature, Ministry of agriculture in Jordan as well as the personal interviews of Sweimeh residents to determine the location of Prosopis and Tamrix in the village in 1999. Then we calculated the total area of IPS for the 1999.

Figure 5. If 70% is the data for 2017, how was % of invaded areas in 1999-2013 calculated?

Our response: We used the data in July to reduce the amount of error. No agricultural activities was occurred at that time of the year. So only trees are available. Also, we conducted interviews with the residents of the area and used the historical data from Royal Society for the Conservation of Nature, Ministry of agriculture in Jordan to estimate the total area of Prosopis in 1999. Because the total area of Prosopis in 1999 was about 60% and 70% for 2017 (about 10% change). We assumed that the % of IPS is 65% for the 2003, 2006 and 2013. We clarified this point in the revised version of the manuscript.

In section 3.2:

Lines 239-243 and 262-271 are not results, they should be in the introduction

Our response: We agree. Lines 239-243 and 262-271 removed from the result section.

Lines 274-313 are not results of this study but of other studies.

Our response: We agree. Lines 274-313 removed from the result section.

What are the data of Table 4 used for? Which conclusions are obtained from them? Could these data be combined with image classifications results, for example, to estimate IPS biomass?

Our response: Good point. Table 4 combined with Figure 5 to show the total wood production of Prosopis tree.

Reviewer 2 Report

The manuscript entitled „Evaluating the effect of Prosopis juliflora, an alien invasive species, on land cover change using remote sensing approach” deals with the spread of the tree Prosopis in a part of the Jordan Valley (Jordan) over a period of time of roughly two decades, i.e. between 1999 and 2017. Employing Landsat ETM+ satellite imagery of five time slices (1999, 2003, 2006, 2013 and 2017), the changes in coverage of four broad land cover classes (soil, urban, vegetation and water) are analyzed. Monitoring invasive alien plant species is a highly relevant topic, and employing remote sensing techniques allows to expand areal coverage, so the contribution in general is highly welcome. However, I see several points which should be addressed by the authors prior to publication by a major revision.

General remarks

L26 (and several other places throughout the manuscript): It does not become sufficiently clear whether the class “vegetation” also include other species or just Prosopis? Only in Line 218-220 we are informed that ~30% of vegetation cover is provided by other species. What are those species, to what life-form do they belong (trees, grasses, annual species, etc.) and how can they be separated from Prosopis in the remote sensing analyses? Are the other species still present in July or are these annual species which are already dormant at this point? The treatment of the class “Vegetation” must be better explained.

L34-35: that is an important point which should be picked up in the Discussion and/or Conclusion chapter

L142: Results and Discussion: I recommend to dived this chapter into 3. Results (i.e. own results of the analyses performed) and 4. Discussion. Into the Discussion chapter the following paragraphs should be moved: L62-67 (For the Introduction general assessments (including references) are sufficient); L238-245; L287-297; also, the Chapter 3.3 Prosopis benefits and costs should be moved to Discussion.

L 156-157: Figure 2: There is a very obvious shift of the class “vegetation” from a distribution mainly near the coast in 1999 and away from coast after 2006, what is the reason for that? And how is the much higher cover of the class “Urban” in 1999 compared to the later images to be explained?

L168-182: this is not Results, move either to Introduction, Methods or to the Discussion chapter.

Specific remarks:

L25: how is the decrease of urban (11%) related to Prosopis?

L29: instead of invasion better use cover/coverage

L39: better use societal instead of social

L42: negative impacts: what about N-Enrichment in the soils? Many Fabaceae do this, Prosopis as well???

L61: Invasive Plants: I recommend a consequent use of abbreviations IAS and IPS, better yet just one of the two as they have more or less the same meaning (also L75).

L69: What was the reason for the Introduction of Prosopis to the Jordan Valley by the Ministry of Agriculture?

L78: “physiology and structure of vegetation”: Is the resolution of Landsat able to capture this? Please also give some hints on the limitations of Landsat either in the introduction or the Methods chapter.

L97: impact of Prosopis: actually, you are evaluating the spread of Prosopis not its impact

L104-105: it would be nice to have a climate chart included to get an idea of seasonal changes of temperature and precipitation. The chart could be included in Figure 1

L111: pre-processing: Did you perform a radiometric normalization of the different Landsat images? This is a procedure which is actually required prior to a meaningful change detection? Please explain. Also, the limitations of Landsat data should shortly be discussed (see comment above)

L127-128: I am wondering whether one day is enough for hundred 10.000 square meter plots?

L144: this is Introduction

L218: what are the native plants present in summer?

L238: grasslands: this is not distinguished in the Landsat classification, there is just one class "vegetation"

L247-248: “The total invasion percentage of Prosopis compared to non-invaded area was 6% 1n 1999, 13% in 2003, 20% in 2006, 22% in 2013 and 23% in 2017: these are the values for non-invaded land in Figure 5, please clarify (also L272)

L252-254: redundant information, already mentioned in Study Area section

L275: …warm and humid…: Jordan valley with 50 mm annual precipitation is far from humid! Please clarify

L288: lack instead of lake

L308: invasive: do you mean invasion by…?

L309: general: do you mean generate…?

L312: “…supporting local use of invasive alien species…”: how? For fire wood, fodder, etc. Please give some examples

L322: laws instead of lows

L332-349: Chapter 4. Conclusions is redundant and can be removed

Author Response

Thank you for the constructive and thoughtful comments. We have revised the manuscript comments and suggestions. The revised text in the new version of the manuscript is in red color. As outlined below, we addressed all your concern points.  We feel the revisions incorporated in the manuscript have strengthened the manuscript and hope you will now find it suitable for publication.

The manuscript entitled „Evaluating the effect of Prosopis juliflora, an alien invasive species, on land cover change using remote sensing approach” deals with the spread of the tree Prosopis in a part of the Jordan Valley (Jordan) over a period of time of roughly two decades, i.e. between 1999 and 2017. Employing Landsat ETM+ satellite imagery of five time slices (1999, 2003, 2006, 2013 and 2017), the changes in coverage of four broad land cover classes (soil, urban, vegetation and water) are analyzed. Monitoring invasive alien plant species is a highly relevant topic, and employing remote sensing techniques allows to expand areal coverage, so the contribution in general is highly welcome. However, I see several points which should be addressed by the authors prior to publication by a major revision.

 General remarks

L26 (and several other places throughout the manuscript): It does not become sufficiently clear whether the class “vegetation” also include other species or just Prosopis? Only in Line 218-220 we are informed that ~30% of vegetation cover is provided by other species. What are those species, to what life-form do they belong (trees, grasses, annual species, etc.) and how can they be separated from Prosopis in the remote sensing analyses? Are the other species still present in July or are these annual species which are already dormant at this point? The treatment of the class “Vegetation” must be better explained.

Our response: We agree. The ‘vegetation’ class was better explained in the revised version of the manuscript.

L34-35: that is an important point which should be picked up in the Discussion and/or Conclusion chapter.

Our response: We agree. This point discussed in the discussion section of the revised manuscript.

L142: Results and Discussion: I recommend to dived this chapter into 3. Results (i.e. own results of the analyses performed) and 4. Discussion. Into the Discussion chapter the following paragraphs should be moved: L62-67 (For the Introduction general assessments (including references) are sufficient); L238-245; L287-297; also, the Chapter 3.3 Prosopis benefits and costs should be moved to Discussion.

Our response: We agree. Results and Discussion was separated to two sections in the revised version of the manuscript.

L 156-157: Figure 2: There is a very obvious shift of the class “vegetation” from a distribution mainly near the coast in 1999 and away from coast after 2006, what is the reason for that? And how is the much higher cover of the class “Urban” in 1999 compared to the later images to be explained?

Our response: We agree. We attributed the reduction in ‘Urban’ class to high demand for water and the reduction in governmental subsidies for agriculture in the last two decades which leaded farmers to abandon their lands. We clarified this point in the revised version (discussion section) of the manuscript.

L168-182: this is not Results, move either to Introduction, Methods or to the Discussion chapter.

 Our response: We agree. Lines 168-182 has been removed from the results section in the revised version of the manuscript.

Specific remarks:

L25: how is the decrease of urban (11%) related to Prosopis?

Our response: Good point. We clarified this point in the discussion section of the revised manuscript.

L29: instead of invasion better use cover/coverage

Our response: ‘invasion’ replaced by ‘coverage’

L39: better use societal instead of social

Our response: ‘social’ replaced by ‘societal’

L42: negative impacts: what about N-Enrichment in the soils? Many Fabaceae do this, Prosopisas well???

Our response: We agree. Text revised.

L61: Invasive Plants: I recommend a consequent use of abbreviations IAS and IPS, better yet just one of the two as they have more or less the same meaning (also L75).

Our response: We agree. ‘IPS’ was used across the revised version of the manuscript. However, we used ‘IAS’ in the revised version because the term include animals.

L69: What was the reason for the Introduction of Prosopis to the Jordan Valley by the Ministry of Agriculture?

Our response: The reason was for afforestation programs. Text revised.  

L78: “physiology and structure of vegetation”: Is the resolution of Landsat able to capture this? Please also give some hints on the limitations of Landsat either in the introduction or the Methods chapter.

Our response: Text revised. In addition, the limitation of Landsat images added to the revised section of introduction as follow: “Although Landsat images are recommended for monitoring land use/land cover studies (Willis, 2015), the spatial resolution (30 m) and the saturation of the optical signal at high biomass density and cloud cover can significantly impact their use; including the phenological information useful for discriminating some vegetation types [21]. Therefore, fusion of satellite and ancillary data may improve the accuracy of remotely-sensed data from Landsat [16,21].”

L97: impact of Prosopis: actually, you are evaluating the spread of Prosopis not its impact

Our response: We agree. ‘impact’ replaced by ‘spread’. 

L104-105: it would be nice to have a climate chart included to get an idea of seasonal changes of temperature and precipitation. The chart could be included in Figure 1

Our response: We agree. Climate chart added to Figure 1. 

L111: pre-processing: Did you perform a radiometric normalization of the different Landsat images? This is a procedure which is actually required prior to a meaningful change detection? Please explain. Also, the limitations of Landsat data should shortly be discussed (see comment above)

Our response: Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes logarithm in ENVI was used for radiometric normalization of the different Landsat images. Text revised.

In addition, the limitation of Landsat data was discussed in the revised version of the manuscript.  

L127-128: I am wondering whether one day is enough for hundred 10.000 square meter plots?

Our response: Good point. A multiple ground survey visits were conducted; started from July 15 to August 20, 2017. Text revised.  

L144: this is Introduction

Our response: We agree. Text revised.

L218: what are the native plants present in summer?

Our response: We agree. Text revised as follow: ‘Therefore, the vegetation cover in the images represented mostly the non-agricultural crops, IPS (Prosopis) and sparse native plants such as, Juncus, Phoenix and Tamarix.’

L238: grasslands: this is not distinguished in the Landsat classification, there is just one class "vegetation"

Our response: We agree. Line 238 deleted.

L247-248: “The total invasion percentage of Prosopis compared to non-invaded area was 6% 1n 1999, 13% in 2003, 20% in 2006, 22% in 2013 and 23% in 2017: these are the values for non-invaded land in Figure 5, please clarify (also L272)

Our response: The percentage for the terrestrial land. Text revised as follow: “The total invasion percentage of Prosopis compared to non-invaded land (terrestrial) was 5% 1n 1999, 13% in 2003, 20% in 2006, 22% in 2013 and 23% in 2017.”

L252-254: redundant information, already mentioned in Study Area section

Our response: We agree. Lines 252-254 deleted.

L275: …warm and humid…: Jordan valley with 50 mm annual precipitation is far from humid! Please clarify

Our response: We agree. The word ‘humid’ removed from the revised version of the manuscript. 

L288: lack instead of lake

Our response: ‘lake’ replaced by ‘lack’.

L308: invasive: do you mean invasion by…?

Our response: Correct. Text revised.

L309: general: do you mean generate…?

Our response: Correct. ‘general’ replaced by ‘generate’.

L312: “…supporting local use of invasive alien species…”: how? For fire wood, fodder, etc. Please give some examples

Our response: We agree. Text revised. 

L322: laws instead of lows

Our response: ‘lows’ replaced by ‘laws’.

L332-349: Chapter 4. Conclusions is redundant and can be removed

Our response: We agree. Chapter 4 removed.

Reviewer 3 Report

The paper is well-written and useful. Assessing and predicting range expansion of invasive alien species is pivotal to design addressed management plans. This is particularly evident within and in the surroundings of extreme environments e.g. deserts.

Therefore, I recommend the publish to be published on “Sustainability” and I only have few minor revisions.

1. Why did the accuracy decrease between 1999 and 2017?
2. LINES 39-42. IPS might also provoke an alteration in landcover, with several consequences. See Lazzaro, L., Mazza, G., d'Errico, G., Fabiani, A., Giuliani, C., Inghilesi, A. F., & Roversi, P. F. (2018). How ecosystems change following invasion by Robinia pseudoacacia: Insights from soil chemical properties and soil microbial, nematode, microarthropod and plant communities. Science of the Total Environment, 622, 1509-1518.
3. LINES 73-74. Assessing the distribution of a species is pivotal for its conservation and management. This should be emphasized here.
4. LINE 76. Please, add a reference.
5. LINES 87-99. In the last part of the introduction, I suggest you to clarify aims and predictions, so that discussion would be better explained and showed.
6. DISCUSSION should be better focused on fulfillment/denying of predictions stated at the end of the introduction.

Emiliano Mori (University of Siena)

Author Response

Thank you for the constructive and thoughtful comments. We have revised the manuscript comments and suggestions. The revised text in the new version of the manuscript is in red color. As outlined below, we addressed all your concern points.  We feel the revisions incorporated in the manuscript have strengthened the manuscript and hope you will now find it suitable for publication.

Reviewer comments

The paper is well-written and useful. Assessing and predicting range expansion of invasive alien species is pivotal to design addressed management plans. This is particularly evident within and in the surroundings of extreme environments e.g. deserts.

Therefore, I recommend the publish to be published on “Sustainability” and I only have few minor revisions.

1. Why did the accuracy decrease between 1999 and 2017?

Our response: Good point. We clarify this point in the result section.

1. LINES 39-42. IPS might also provoke an alteration in landcover, with several consequences. See Lazzaro, L., Mazza, G., d'Errico, G., Fabiani, A., Giuliani, C., Inghilesi, A. F., & Roversi, P. F. (2018). How ecosystems change following invasion by Robinia pseudoacacia: Insights from soil chemical properties and soil microbial, nematode, microarthropod and plant communities. Science of the Total Environment, 622, 1509-1518.

Our response: We agree. The following paragraph added the introduction section 

“Lazzaro et al. [5] studied how ecosystems change following invasion by Robinia pseudoacacia. They found that the abundance and richness of microarthropods, richness of nematodes, and richness and diversity of plant communities decreased significantly in invaded stands. In fact, they confirmed that the IPS can transform several ecosystem components, modifying the plant-soil community and affecting biodiversity at different levels.”

1. LINES 73-74. Assessing the distribution of a species is pivotal for its conservation and management. This should be emphasized here.

Our response: We agree. The paragraph revised.

1. LINE 76. Please, add a reference.

Our response: Reference added.

1. LINES 87-99. In the last part of the introduction, I suggest you to clarify aims and predictions, so that discussion would be better explained and showed.
2. DISCUSSION should be better focused on fulfillment/denying of predictions stated at the end of the introduction.

Our response: Our hypothesis in the last of the introduction is that remote sensing has the potential to detect Prosopis cover change across the study period. In the revised version of the manuscript, we confirm and discussed in-depth how these remotely-sensed data were effective.    

Round 2

Reviewer 1 Report

The authors have addressed the reviewer comments

Reviewer 2 Report

Dear Authors,

thank you very much for your efforts to improve your manuscript "Evaluating the effect of Prosopis juliflora, an alien invasive species, on land cover change using remote sensing approach". From my point of view you have nicely addressed my concerns made during the first round of review in the revised version. I believe that the manuscript significantly improved and now warrants publication in MDPI Sustainability. I found two small mistakes, which should be corrected: in L235 it must read in instead of 1n, and the Discussion should go as Chapter 4 and the subchapters have to be numbered accordingly.