Effects of Urbanization-Induced Land Use Changes on Ecosystem Services: A Case Study of the Anhui Province, China

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study provides a comprehensive analysis of how urbanization drivers interact with land use and land cover (LULC) changes to influence ecosystem services (ESs) in Anhui Province, China. The integration of spatial analysis, machine learning (random forest), and structural modeling (PLS-PM) offers robust insights into the spatiotemporal dynamics of ESs and their linkages to urbanization. The findings highlight critical policy implications for balancing urban development with ecological sustainability. While the methodology is sound and the conclusions are well-supported, some areas require clarification or expansion to strengthen the study’s contribution.

Model Validation and Performance:

The RF model exhibits low R² values for grassland and barren land (mentioned in Section 3.3.1). The authors should discuss whether these limitations affect the overall conclusions, especially given the minor proportion of these LULC types in Anhui.

Exclusion of Cultural Services:

While justified, the omission of cultural services (e.g., recreation, aesthetic value) narrows the ES assessment. A brief discussion on how this exclusion might bias the results would strengthen the limitations section.

Data Resolution and Scale:

The 30m resolution is appropriate for provincial-scale analysis, but the authors should acknowledge whether finer resolutions (e.g., 10m) could improve ES assessments, particularly in heterogeneous urban areas.

Generalizability:

The findings are context-specific to Anhui’s socio-ecological landscape. The conclusion should explicitly caution against direct extrapolation to other regions without considering local drivers (e.g., policy frameworks, biophysical conditions).

Policy Recommendations:

While the recommendations are practical, they lack specificity. For example, how might "mixed-use urban development" be implemented in Anhui’s central urbanizing zones?

Figures and Tables:

Figure 1b (LULC transition) and Figure 5 (hotspot analysis) are critical but lack sufficient labeling. Ensure legends, scales, and temporal markers (e.g., years) are clearly visible.

Table 3 (urbanization drivers) should include units for clarity (e.g., GDP in RMB/year).

Author Response

We sincerely thank you for the valuable and insightful feedback. Your suggestions have greatly contributed to improving the clarity, depth, and policy relevance of our manuscript. Below, we provide point-by-point responses to each comment and outline the corresponding revisions in the manuscript. The reviewer comments are laid out below in italicized font and specific concerns have been numbered. Our response is given in normal font and changes or additions to the manuscript are given in the blue text. The revisions related to language polishing have been marked in red in the original manuscript.

 

1. Comment: Model Validation and Performance

The RF model exhibits low R² values for grassland and barren land (mentioned in Section 3.3.1). The authors should discuss whether these limitations affect the overall conclusions, especially given the minor proportion of these LULC types in Anhui.

Response:

Thank you for pointing this out. We have added a brief discussion in Discussion to acknowledge that the relatively low R² values for grassland and barren land are due to their small spatial proportions in Anhui Province and relatively stable land cover over time. We clarified that while these lower values may affect the model’s precision for these specific LULC types, they do not significantly impact the overall conclusions, as the main ES changes are driven by forest, cropland, and construction land dynamics.

Revision made in Section 4.3, line 550-557.

2. Comment: Exclusion of Cultural Services

While justified, the omission of cultural services (e.g., recreation, aesthetic value) narrows the ES assessment. A brief discussion on how this exclusion might bias the results would strengthen the limitations section.

Response:

We have added in the Discussion section discussing how the exclusion of cultural services may lead to an underestimation of the total value and multifunctionality of ecosystem services. In this study, we focused on biophysical and regulating services due to data limitations and the challenges of quantifying cultural services consistently at the regional scale. We acknowledged that excluding cultural services may underestimate the total value of ecosystem services in certain areas, especially those with significant cultural or recreational importance.

Revision made in Section 4.3, line 540-545.

3. Comment: Data Resolution and Scale

The 30m resolution is appropriate for provincial-scale analysis, but the authors should acknowledge whether finer resolutions (e.g., 10m) could improve ES assessments, particularly in heterogeneous urban areas.

Response:

We have included in Section 4.4 noting that while the 30m resolution is suitable for provincial-scale analysis, the use of higher-resolution datasets (e.g., 10m Sentinel-2 imagery) could enhance the accuracy of ES assessments in highly heterogeneous landscapes such as dense urban zones. This is suggested as a future research direction.

Revision made in Section 4.4, line 546-550.

4. Comment: Generalizability

The findings are context-specific to Anhui’s socio-ecological landscape. The conclusion should explicitly caution against direct extrapolation to other regions without considering local drivers (e.g., policy frameworks, biophysical conditions).

Response:

We agree that the findings are context-specific to Anhui’s socio-ecological landscape. To address this, we have added a statement in the conclusion to caution against direct extrapolation of the results to other regions without accounting for local drivers such as biophysical settings and policy contexts. 

Revision made in Conclusion, line 596-599.

5. Comment: Policy Recommendations

While the recommendations are practical, they lack specificity. For example, how might ‘mixed-use urban development’ be implemented in Anhui’s central urbanizing zones?

Response:

We have refined the policy recommendation section to include more specific guidance. In particular, we elaborated on how mixed-use urban development could be implemented in central Anhui by promoting land-use zoning. We also specified the renewable resource utilization and adaptive land-use planning in rapidly urbanizing areas of Anhui Province.

Revision made in Section 4.2, line 514-524.

6. Comment: Figures and Tables

Figure 1b (LULC transition) and Figure 5 (hotspot analysis) are critical but lack sufficient labeling. Ensure legends, scales, and temporal markers (e.g., years) are clearly visible.

Response:

We have revised Figures 1b and 5 to improve their clarity by adding legends, scale bars, and year labels. The updated figures now provide more informative visual context.

Revised Figure 1b and Figure 5.

7. Comment: Table 3 (urbanization drivers) should include units for clarity (e.g., GDP in RMB/year).

Response:

We have added units to all variables in Table 3, including specifying GDP as “10⁴ RMB yr-1 ” and other indicators as applicable.

Revised Table 3.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The introduction provides a solid overview of ecosystem services (ES) and urbanisation dynamics in Anhui Province, but would benefit from a clearer link to other frameworks, such as the European framework. For example, integrating the Mapping and Assessment of Ecosystems and Ecosystem Services (MAES) framework would place the work within a widely used methodology for assessing ES at different scales and in different land use contexts. Similarly, citing the 'EU Biodiversity Strategy 2030' would underline the policy relevance of their findings by showing how ambitious European targets for habitat restoration and species protection parallel the sustainability challenges faced by Anhui.

In addition, the manuscript should make explicit reference to key principles of European ecosystem planning to strengthen its methodological justification. The INSPIRE Directive, which requires interoperable spatial data infrastructures across the EU, could be cited to support the high-resolution land cover mapping approach. Similarly, the European Green Deal's goal of 'nature-positive' urban development provides a clear policy perspective for interpreting its hotspot and coldspot analyses: aligning its spatial prioritisation with the Green Deal's objectives highlights the practical utility of its results for decision-makers seeking to balance urban growth with ecological integrity.

The literature review on the interactions between multiple drivers of urbanisation and ESs should be expanded to include more recent studies (10.1016/j.futures.2024.103399; 10.1016/j.ecocom.2009.10.006) on urban-rural trade-offs of ESs.

It is essential that the Methods section includes precise details of the parameterisation of each model to ensure that other researchers can replicate the analyses. For example, for GWR, specify the type of kernel used and the band optimisation criterion (e.g. AICc or CV), as well as the optimal band search range; for Random Forest, specify the number of trees, the maximum depth, the number of variables considered in each partition, and the cross-validation method used to adjust these hyperparameters.

Similarly, when describing K-means clustering, explain the method used to determine the optimal number of clusters (e.g. 'silhouette' or 'elbow' criterion) and the reasons why this partitioning adequately reflects the spatial variability of the ecosystem services and urbanisation indicators in the study.

In the discussion, it would be relevant to link Anhui's current land use plans to an ecosystem planning approach, noting how the province's 'ecological civilisation' strategy promotes the creation of green corridors and protected areas, which could be strengthened by ecosystem planning criteria of connectivity and multifunctionality (e.g. integration of buffer zones and pollination services). It would also be useful to discuss the uncertainties inherent in the input datasets - such as the sensitivity of carbon sequestration or water filtration parameters in InVEST, the accuracy of satellite land cover classification and temporal resolution - and how these limitations may affect the identification of green infrastructure priorities and the allocation of restoration areas in an adaptive landscape management framework.

The discussion on policy implications will be stronger if its recommendations are explicitly linked to existing European instruments. For example, advocating greater protection of forests and wetlands makes direct reference to the EU Nature Restoration Act, while promoting mixed-use green corridors coincides with the Urban Greening Plans promoted by the Covenant of Mayors. Framing your land use management suggestions in this way will demonstrate that your study not only advances academic knowledge, but also offers practical guidance aligned with key ecosystem policy frameworks.

Comments on the Quality of English Language

The manuscript is generally clear. However, some complex sentences could be simplified.

Author Response

We sincerely thank you for the valuable and insightful feedback. Your suggestions have greatly contributed to improving the clarity, depth, and policy relevance of our manuscript. Below, we provide point-by-point responses to each comment and outline the corresponding revisions in the manuscript. The reviewer comments are laid out below in italicized font and specific concerns have been numbered. Our response is given in normal font and changes or additions to the manuscript are given in the green text. The revisions related to language polishing have been marked in red in the original manuscript.

 

1. Comment: Integrating the Mapping and Assessment of Ecosystems and Ecosystem Services (MAES) framework would place the work within a widely used methodology for assessing ES at different scales and in different land use contexts. Similarly, citing the 'EU Biodiversity Strategy 2030' would underline the policy relevance of their findings by showing how ambitious European targets for habitat restoration and species protection parallel the sustainability challenges faced by Anhui.

Response:

We fully agree with the importance of situating our research within globally recognised assessment frameworks. Accordingly, we have revised the introduction to include a discussion of the MAES framework, highlighting its relevance to ecosystem service assessments across different land use contexts and scales. We also introduced the EU Biodiversity Strategy 2030 to emphasise the parallel sustainability challenges between Europe and China, particularly in terms of habitat restoration and ecosystem service protection. These additions help to demonstrate the broader policy significance of our study.

Revision have been added in Introduction, Paragraphs 1, line 37-40.

2. Comment: The European Green Deal's goal of 'nature-positive' urban development provides a clear policy perspective for interpreting its hotspot and coldspot analyses: aligning its spatial prioritisation with the Green Deal's objectives highlights the practical utility of its results for decision-makers seeking to balance urban growth with ecological integrity.

Response:

We totally agree that the idea of “nature-positive” urban development in the European Green Deal provides a useful policy perspective. We have added a short discussion on how our spatial prioritization results can inform decision-makers seeking to reconcile urban growth with ecological sustainability. This helps to underscore the policy relevance and applicability of our study outcomes.

Revision have been added in Section 4.2, line 473-478.

3. Comment: The literature review on the interactions between multiple drivers of urbanisation and ESs should be expanded to include more recent studies (10.1016/j.futures.2024.103399; 10.1016/j.ecocom.2009.10.006) on urban-rural trade-offs of ESs.

Response:

We have expanded the literature review section to include the suggested studies- references[19] and [20] (10.1016/j.futures.2024.103399; 10.1016/j.ecocom.2009.10.006), which offer valuable insights into trade-offs across urban-rural gradients.

Revisions made in Introduction, Paragraphs 3, line 56-59.

4. Comment: It is essential that the Methods section includes precise details of the parameterisation of each model to ensure that other researchers can replicate the analyses. For example, for GWR, specify the type of kernel used and the band optimisation criterion (e.g. AICc or CV), as well as the optimal band search range; for Random Forest, specify the number of trees, the maximum depth, the number of variables considered in each partition, and the cross-validation method used to adjust these hyperparameters.

Response:

We have revised the Methods section to include the missing parameter details for both the GWR and Random Forest models, including kernel type, bandwidth selection method, and hyperparameters used.

Detailed technical specifications added in Section 2.3.2 and 2.5.1., line 153-155, line 240-243.

 

5. Comment: It would be relevant to link Anhui's current land use plans to an ecosystem planning approach, noting how the province's 'ecological civilisation' strategy promotes the creation of green corridors and protected areas, which could be strengthened by ecosystem planning criteria of connectivity and multifunctionality (e.g. integration of buffer zones and pollination services).

Response:

We have added a paragraph in the discussion section highlighting how our findings can support the province’s efforts in establishing green corridors and protected areas. Specifically, we emphasize the role of ecosystem planning principles, such as spatial connectivity and multifunctionality in strengthening these initiatives, including the integration of buffer zones and ecosystem service functions like pollination.

Revisions made in Section 4.2, line 503-509.

6. Comment: It would also be useful to discuss the uncertainties inherent in the input datasets - such as the sensitivity of carbon sequestration or water filtration parameters in InVEST, the accuracy of satellite land cover classification and temporal resolution - and how these limitations may affect the identification of green infrastructure priorities and the allocation of restoration areas in an adaptive landscape management framework.

Response:

We have expanded the limitations section to discuss the uncertainties associated with carbon and water-related parameters in InVEST, as well as the potential errors from satellite-based land cover classification and the temporal mismatch between datasets. We also briefly reflect on how these uncertainties might influence spatial planning outcomes and stress the need for adaptive, iterative landscape management.

Revisions made in Section 4.3, line 557-566.

7. Comment: The discussion on policy implications will be stronger if its recommendations are explicitly linked to existing European instruments. For example, advocating greater protection of forests and wetlands makes direct reference to the EU Nature Restoration Act, while promoting mixed-use green corridors coincides with the Urban Greening Plans promoted by the Covenant of Mayors. Framing your land use management suggestions in this way will demonstrate that your study not only advances academic knowledge, but also offers practical guidance aligned with key ecosystem policy frameworks.

Response:

We have revised the policy implication section to explicitly align our recommendations with relevant European instruments, such as the EU Nature Restoration Act and the Urban Greening Plans under the Covenant of Mayors, thereby enhancing the applicability and translatability of our findings within broader ecosystem governance frameworks.

Revisions made in Section 4.2, line 524-535.

  8. Comment:The manuscript is generally clear. However, some complex sentences could be simplified.

We have reviewed the manuscript thoroughly and revised several long or complex sentences to improve clarity and readability. These changes have been made throughout the text, particularly in the abstract, introduction, and discussion sections.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In this article, the authors presented an analysis of Ecosystem Services (ES) in a Chinese region through a geospatial model. The manuscript is interesting enough to keep reading it to the end, and has scientific soundness. Also, the case study would be interesting for the audience. Minor suggestions for the authors are:

1. Please change the currency from RMB to dollars to have an idea of how rich or poor was Anhui Province at the time of the study, which I assume is 2020 (line 88).
2. Please, review the English language in the entire manuscript because I found several grammatical errors.
3. Please homogenize the nomenclature in figure 1: all LULC o all names (land use/land cover)
4. Please reference that the layers or data used to create the urbanization drivers in Table 1 will be found in Table 3. 
5. Figure 6 it is not clear, please explore other ways to present this information.
6. Table 4 could be a supplementary material.

Author Response

We sincerely thank you for the valuable and insightful feedback. Your suggestions have greatly contributed to improving the clarity, depth, and policy relevance of our manuscript. Below, we provide point-by-point responses to each comment and outline the corresponding revisions in the manuscript. The reviewer comments are laid out below in italicized font and specific concerns have been numbered. Our response is given in normal font and changes or additions to the manuscript are given in the yellow text. The revisions related to language polishing have been marked in red in the original manuscript.

1. Comment:Please change the currency from RMB to dollars to have an idea of how rich or poor was Anhui Province at the time of the study, which I assume is 2020 (line 88).

Response:

 We have converted all economic indicators from RMB to USD using the average exchange rate in 2000 (1 USD ≈ 8.30 RMB) and 2020 (1 USD ≈ 6.90 RMB). This helps provide a clearer understanding of the economic context of Anhui Province at the time of the study. The corresponding changes have been made in the main text (line 92 and throughout where applicable).

2. Comment:Please review the English language in the entire manuscript because I found several grammatical errors.

Response:

We appreciate your attention to language quality. The entire manuscript has been thoroughly revised by a native English speaker and further checked using professional language editing tools to correct grammar, sentence structure, and improve clarity.

3. Comment:Please homogenize the nomenclature in figure 1: all LULC or all names (land use/land cover).

Response:

We have revised Figure 1 to use consistent terminology throughout. All references now uniformly use “LULC” on the context.

4. Comment:Please reference that the layers or data used to create the urbanization drivers in Table 1 will be found in Table 3.

Response:

A note has been added in the main text to clarify that the detailed urbanization drivers are listed in Table 3 (line 110).

5. Comment:Figure 6 it is not clear, please explore other ways to present this information.

Response:

we have replaced the original bar chart in Figure 6 with a heatmap to present the information more clearly and intuitively. We believe this new visualization enhances the readability and facilitates better comparison among the data.

6. Comment:

Table 4 could be a supplementary material.

Response:

Table 4 has been moved to the supplementary material section to streamline the main text and maintain focus on the core findings. A reference to this supplementary table has been added in the appropriate place in the main manuscript.

Author Response File: Author Response.pdf