Identifying Landscape Characteristics That Maximize Ecosystem Services Provision

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General comments

Your work is well motivated, the structure is somewhat appropriate, and the manuscript is not missing any key details. The methods used are appropriate for the objectives of the work and, the resulting figures are sufficient, informative and of good quality helping to follow the reasoning throughout the manuscript. Materials and methods section needs some minor adjustments in order to make more understandable the data used in the analyses. In this section and in the successive ones, references to the use of proxies should be included (see specific comments).

 

Specific comments

 Lines 23-38. As reported in the Authors guideline, the abstract should be a single paragraph and should follow the style of structured abstracts, but without headings. You should remove the headings (Background:, Methods:, …).

Line 27. “GAM”. When defined for the first time, the acronym/abbreviation/initialism should be added in parentheses after the written-out form.

Line 60. The “ES” acronym (ecosystem services) should be used before (line 47-47).

Lines 60 13. This sentence doesn’t sound very well; it should be reformulated.

Lines 69-72. This sentence doesn’t sound very well; it should be reformulated.

Line 118. You can use only the acronym ES.

Line 120. After “Benedetti et al. (2020)”, put a dot.

Line 120. After “Benedetti et al. (2020)”, you should indicate the corresponding reference number (probably is 9).

Line 120. After “Munafò (2018)”, you should indicate the corresponding reference number (probably is 49).

Lines 121-123. This statement needs a reference.

Lines 118-125. You indicated that proxies are used to map the ES; you should indicate in a better way the procedures and resources used. Moreover, here and in the results ad discussion sections, you should refer to the problem of this approach and provide an estimate of the error.

Line 136. After “Benedetti et al. (2020)” you should indicate the corresponding reference number (probably is 9).

Line 210. Istructions for authors indicate that “All Figures, Schemes and Tables should be inserted into the main text close to their first citation and must be numbered following their number of appearance”; delete the “3.2 Figures, Tables and Schemes” paragraph and insert tables and figures in the proper part of the manuscript.

Line 231. The explored ecosystem services are four, not “six” as indicated in the table caption. Adjust the caption.

Line 239. Change “six” in four.

Author Response

Reviewer 1

General comments
Your work is well motivated, the structure is somewhat appropriate, and the manuscript is not missing any key details. The methods used are appropriate for the objectives of the work and, the resulting figures are sufficient, informative and of good quality helping to follow the reasoning throughout the manuscript. Materials and methods section needs some minor adjustments in order to make more understandable the data used in the analyses. In this section and in the successive ones, references to the use of proxies should be included (see specific comments).

Response: We deeply thank the reviewer for recognizing the importance of our work and for taking the time to thoroughly revise our MS. We did our best to incorporate all comments and suggestions, which greatly improved the quality of our study. 

Specific comments
Lines 23-38. As reported in the Authors guideline, the abstract should be a single paragraph and should follow the style of structured abstracts, but without headings. You should remove the headings (Background:, Methods:, …).

Response: Thank you, we have removed the headings.

Line 27. "GAM". When defined for the first time, the acronym/abbreviation/initialism should be added in parentheses after the written-out form.

Response: Thank you. We have added the written-out form and GAM between parentheses.

Line 60. The "ES" acronym (ecosystem services) should be used before (line 47-47).

Response: Thank you. We changed the acronym position. 

Lines 60 13. This sentence doesn't sound very well; it should be reformulated.

Response: Thank you. We have reformulated the sentence to improve the clarity. 

Lines 69-72. This sentence doesn't sound very well; it should be reformulated.

Response: Thank you. We have reformulated the sentence to improve the clarity. 

Line 118. You can use only the acronym ES.

Response: Thank you. We have included the acronym ES in the manuscript. 

Line 120. After “Benedetti et al. (2020)”, put a dot.

Response: Done.                               

Line 120. After "Benedetti et al. (2020)", you should indicate the corresponding reference number (probably is 9).

Response: Done

Line 120. After "Munafò (2018)", you should indicate the corresponding reference number (probably is 49).

Response: Done

Lines 121-123. This statement needs a reference.

Response: Done

Lines 118-125. You indicated that proxies are used to map the ES; you should indicate in a better way the procedures and resources used. Moreover, here and in the results ad discussion sections, you should refer to the problem of this approach and provide an estimate of the error.

Response: Following the reviewer's suggestion, we have expanded the methods section to include more detailed information about the procedures and resources used to estimate ES proxies. Please, see lines 113-138.

Estimating the margin of error for this type of data is quite challenging. However, it was accomplished for the cartographic base represented by the land use and cover map in the article by De Fioravante et al. (2022). The main challenge in assessing ecosystem services at the national scale is in the required generalization of the data due to the vastness of the study area. Instead, more accurate data could be found depending on the region, or specific sampling could be conducted on a smaller scale. This would allow for a more nuanced consideration of the varying influence of factors that might affect the estimation of ecosystem services (e.g., climate, soil composition, specific vegetation types). Briefly, we mentioned this in the discussion section as follows:

" 4.2 Limitations of the approach of this study

Estimating the margin of error for each ES data used in this study is quite challenging. However, it was accomplished for the cartographic base represented by the land use and cover map in the article by De Fioravante et al. [51]. Due to the vastness of the study area, the main challenge in assessing ecosystem services at the national scale is the required generalization of the data. Instead, more accurate data could be found depending on the region, or specific sampling could be conducted on a smaller scale. This would allow for a more nuanced consideration of the varying influence of factors that might affect the estimation of ecosystem services (e.g., climate, soil composition, specific vegetation types)."

Cited references

De Fioravante, P., Strollo, A., Assennato, F., Marinosci, I., Congedo, L., and Munafò, M. (2022). High Resolution Land Cover Integrating Copernicus Products: A 2012–2020 Map of Italy. Land 11, 35. doi: 10.3390/land11010035

Line 136. After "Benedetti et al. (2020)" you should indicate the corresponding reference number (probably is 9).

Response: Done

Line 210. Istructions for authors indicate that "All Figures, Schemes and Tables should be inserted into the main text close to their first citation and must be numbered following their number of appearance"; delete the "3.2 Figures, Tables and Schemes" paragraph and insert tables and figures in the proper part of the manuscript.

Response: Thank you very much for the indication. We have removed the section and inserted figures and tables in correspondence to the text.

Line 231. The explored ecosystem services are four, not "six" as indicated in the table caption. Adjust the caption.

Response: Done

Line 239. Change "six" in four.

Response: Done

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Page 5, Line - 204. The authors did not sufficiently justify the need to use the Shannon index, because it was possible to use the Pelau index. Which is essentially a normalization of the Shannon index.

Page 5, Line 205-207. It is not clear how the authors calculated the "confidence intervals". It is known that in practice the calculation of confidence intervals and hypothesis testing is usually performed using Fisher's transformation.

Author Response

Reviewer 2

General comment: Thank you very much for your valuable comments and suggestions. We have carefully considered the reviewer's input, which will allow us to improve our manuscript. Please see below our point-by-point responses:

Question 1

Page 5, Line - 204. The authors did not sufficiently justify the need to use the Shannon index, because it was possible to use the Pelau index. Which is essentially a normalization of the Shannon index.

Response 1: We recognize that using the Shannon diversity index has certain limitations. The index can vary depending on the data type and ranges from 0 to infinity. One of the limitations is its potential negative bias at small sample sizes, as noted by Konopiński in 2020. To address this, we utilized a large sample size of 308,564 observations to minimize any bias in our estimation of the Shannon index. Additionally, for the GAM models used in our study, standardizing predictors, such as the Shannon index, is not strictly necessary (Wood, 2017). Therefore, we used the Shannon index and weighted edge density to describe landscape heterogeneity, focusing on landscape composition (e.g., evenness or Shannon diversity index) and configuration (edge density). We also considered weighted edge density as an additional proxy of landscape heterogeneity. We chose the Shannon index due to its common usage in ecological studies (Tonetti et al., 2023) and its comparability and understandability for readers. This index is widely used in ecological literature (see below some examples):

Benedetti, Y., Morelli, F., Munafò, M., Assennato, F., Strollo, A., and Santolini, R. (2020). Spatial associations among avian diversity, regulating and provisioning ecosystem services in Italy. Ecol. Indic. 108, 105742. doi: 10.1016/j.ecolind.2019.105742

Burel, F., Baudry, J., Butet, A., Clergeau, P., Delettre, Y., Le Coeur, D., et al. (1998). Comparative biodiversity along a gradient of agricultural landscapes. Acta Oecologica 19, 47–60. doi: 10.1016/S1146-609X(98)80007-6

Cadotte, M. W. M. ., Jonathan Davies, T., Regetz, J., Kembel, S. W. S. W., Cleland, E., Oakley, T. H., et al. (2010). Phylogenetic diversity metrics for ecological communities: Integrating species richness, abundance, and evolutionary history. Ecol. Lett. 13, 96–105. doi: 10.1111/j.1461-0248.2009.01405.x

Castro-Caro, J.C., Barrio, I.C., Tortosa, F.S., 2015. Effects of hedges and herbaceous cover on passerine communities in Mediterranean olive groves. Acta Ornithol. 50, 180–192. Doi: 10.3161/00016454AO2015.50.2.006

Hermy, M., and Cornelis, J. (2000). Towards a monitoring method and a number of multifaceted and hierarchical biodiversity indicators for urban and suburban parks. Landsc. Urban Plan. 49, 149–162. Doi: 10.1016/S0169-2046(00)00061-X 

Johst, K., and Huth, A. (2005). Testing the intermediate disturbance hypothesis: When will there be two peaks of diversity? Divers. Distrib. 11, 111–120. doi: 10.1111/j.1366-9516.2005.00133.x

Kisel, Y., McInnes, L., Toomey, N.H., Orme, C.D.L., 2011. How diversification rates and diversity limits combine to create large-scale species-area relationships. Philos. Trans. R. Soc. London B - Biol. Sci. 366, 2514–2525. Doi: 10.1098/rstb.2011.0022

Lee, M.-B., and Martin, J. A. (2017). Avian Species and Functional Diversity in Agricultural Landscapes: Does Landscape Heterogeneity Matter? PLoS One 12. doi: 10.1371/

Morelli, F., 2013. Relative importance of marginal vegetation (shrubs, hedgerows, isolated trees) surrogate of HNV farmland for bird species distribution in Central Italy. Ecol. Eng. 57, 261–266. Doi: 10.1016/j.ecoleng.2013.04.043

Morelli, F., Pruscini, F., Santolini, R., Perna, P., Benedetti, Y., Sisti, D., 2013. Landscape heterogeneity metrics as indicators of bird diversity: Determining the optimal spatial scales in different landscapes. Ecol. Indic. 34, 372–379. Doi:10.1016/j.ecolind.2013.05.021

Schindler, S., Poirazidis, K., Wrbka, T., 2008. Towards a core set of landscape metrics for biodiversity assessments: A case study from Dadia National Park, Greece. Ecol. Indic. 8, 502–514. Doi: 10.1016/j.ecolind.2007.06.001

Schindler, S., von Wehrden, H., Poirazidis, K., Wrbka, T., Kati, V., 2013. Multiscale performance of landscape metrics as indicators of species richness of plants, insects and vertebrates. Ecol. Indic. 31, 41–48. Doi: 10.1016/j.ecolind.2012.04.012

Stein, A., Gerstner, K., Kreft, H., 2014. Environmental heterogeneity as a universal driver of species richness across taxa, biomes and spatial scales. Ecol. Lett. 17, 866–880. Doi: 10.1111/ele.12277

Xu, C., Huang, Z.Y.X., Chi, T., Chen, B.J.W., Zhang, M., Liu, M., 2014. Can local landscape attributes explain species richness patterns at macroecological scales? Glob. Ecol. Biogeogr. 23, 436–445. Doi: 10.1111/geb.12108

To increase the clarity of our methods, we have now highlighted that: "From the several landscape metrics available for ecological studies [61,62], we chose two metrics, trying to follow a cost-effective strategy. The two metrics used describe two important landscape characteristics, examining mainly aspects of landscape composition (e.g., evenness or Shannon diversity index) and configuration (weighted edge density). Previous studies have focused on both metrics and shown potential biodiversity surrogacy [63,64]. The Shannon diversity is a good indicator of the land cover diversity. It is potentially associated with the overall biodiversity due to the number of available habitats [65,66]. While the weighted edge density is strongly associated with the presence of linear elements, such as hedgerows, tree lines, etc., increasing the overall landscape heterogeneity, the border effect, and potentially significantly affecting biodiversity [64,67]."

Cited references

Wood, S.N. (2017). Generalized Additive Models: An Introduction with R (2nd edition). Chapman and Hall/CRC

Konopiński, M. K. (2020). Shannon diversity index: A call to replace the original Shannon's formula with unbiased estimator in the population genetics studies. PeerJ 8, e9391. doi:10.7717/peerj.9391

Tonetti V et al. (2023). Landscape heterogeneity: concepts, quantification, challenges and future perspectives. Environmental Conservation page 1 of 10. doi: 10.1017/S0376892923000097

Question 2

Page 5, Line 205-207. It is not clear how the authors calculated the "confidence intervals". It is known that in practice the calculation of confidence intervals and hypothesis testing is usually performed using Fisher's transformation.

Response 2:

We calculated the confidence intervals using the standard errors of the model terms, accounting for the uncertainty in the smoothing parameter estimates described by Marra and Wood (2012). We added this reference for clarity as follows:

"The models were displayed graphically, with their 95% confidence intervals accounting for the uncertainty in smoothing parameter estimates [73]."

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript explored the relationship between landscape characteristics and ecosystem services provision, and the conclusions were useful for policy makers for land use planning and management. However, there are some major limitations listed below:

1.       The manuscript estimated four types of ecosystem services such as agriculture production, timber production, but the accuracy of the four were not validated in Italy.

2.       The relationships among the four ecosystem services were analyzed based on correlation analysis. But the significance of the relationship were not given, and partial correlation analysis would be superior to simple correlation analysis due to the intrinsic interactions among these ecosystem services.

3.       Why choose the three kind of landscape variables was not clearly explained, and the variable like patch density/number, shape index, landscape integrity connectivity and so on were all have a critical impact on ecosystem services.

4.       What does the meaning of ‘3.2 Figures, Tables and Schemes’ in lime 210 on page 5.

5.       Is it ‘four’ in line 239 on page 7?

Comments on the Quality of English Language

Few English expressions need to be improved in the introduction part.

Author Response

Reviewer 3

General comment:

The manuscript explored the relationship between landscape characteristics and ecosystem services provision, and the conclusions were useful for policy makers for land use planning and management. However, there are some major limitations listed below:

Answer: Thank you for your valuable comments and suggestions. We have carefully considered the reviewer's inputs, as they will help us enhance our manuscript. Please see our point-by-point responses as follows:

Question 1

1. The manuscript estimated four types of ecosystem services such as agriculture production, timber production, but the accuracy of the four were not validated in Italy.

Response 1: Thank you for the reviewer's comment. The methods used in this study are based on analyzing existing and well-established data at national and international levels. This includes using a validated land use and land cover map, which can be found at De Fioravante et al., 2022. Additionally, to assess carbon storage, we utilized the organic carbon map from the Global Soil Organic Carbon Map (GSOCmap) (FAO and ITPS, 2018) and the National Inventory of Forests and Forest Carbon Stocks (https://www.inventarioforestale.org/it/). For crop and timber production, we relied on the Average Agricultural Values, which are official data from the Italian Revenue Agency  (https://www.agenziaentrate.gov.it/portale/web/guest/schede/fabbricatiterreni/omi/banche-dati/valori-agricoli-medi).

Cited references

De Fioravante, P., Strollo, A., Assennato, F., Marinosci, I., Congedo, L., and Munafò, M. (2022). High Resolution Land Cover Integrating Copernicus Products: A 2012–2020 Map of Italy. Land 11, 35. doi: 10.3390/land11010035

FAO, and ITPS (2018). Global Soil Organic Carbon Map (GSOCmap). Rome.

Question 2

2. The relationships among the four ecosystem services were analyzed based on correlation analysis. However, the significance of the relationship were not given, and partial correlation analysis would be superior to simple correlation analysis due to the intrinsic interactions among these ecosystem services.

Response 2: Thank you to the reviewer for the helpful recommendation, which has improved the quality of our manuscript. Following your suggestion, we analyzed trade-offs and synergies among ES in Italy by calculating Partial correlation analysis using the ppcor R package [75]. Accordingly, the new correlation analysis was included in the methods section. The new results about trade-offs and synergies were updated in the figure, results, and discussion sections as follows:

Methods section "The study examined the trade-offs and synergistic interactions between different ecosystem services in Italy by assessing their spatial congruence. Thus, a partial correlation analysis approach was applied using the R package ppcor [75] to determine if paired ecosystem services were positively or negatively associated, with a significance level of α at 0.05. Consequently, a positive correlation between the two ecosystem services indicated a synergistic effect, while a negative correlation suggested a trade-off relationship."

Results section "All measured ES demonstrated trade-offs or synergic spatial associations (Figure 2). The trade-offs are represented mainly by agricultural production, showing negative coefficient correlations with timber production showing a weak but significant correlation (Pearson correlation r = -0.018, Figure 2), and habitat quality displaying a stronger correlation (Pearson correlation r = -0.449, Figure 2). The synergies were found in the rest of ES. Higher and positive correlation coefficients were found between carbon storage and habitat quality (Pearson correlation r = -0.614, Figure 2). Following this, high and positive correlation values were found between habitat quality and carbon storage (Pearson correlation r = 0.511, Figure 2), and finally, even if weak but significant synergy was found between carbon storage and agricultural production (Pearson correlation r = 0.093, Figure 2)."

Discussion section "It was found that there is a high degree of spatial congruence among various ES, including habitat quality, carbon storage, and timber production, as indicated by positive and significantly higher correlations. The first implication of these results is that conserving a particular ecosystem service (e.g., regulatory one) could also protect the rest of ES, indicating cost-effective conservation planning. On the other hand, the spatial mismatch between agricultural production with timber production, and habitat quality is reflected in the negative associations found. Moreover, agricultural production showed the lowest correlations with habitat quality. These results are similar to a study conducted by Crouzat et al. [78], which found that agricultural production was not bundled or overlapped with the other ES examined. Meanwhile, the rest of the ES studied were bundled or partially overlapped [78]."

Question 3 

3. Why choose the three kind of landscape variables was not clearly explained, and the variable like patch density/number, shape index, landscape integrity connectivity and so on were all have a critical impact on ecosystem services.

Response 3: Thank you very much for your observation. We agree with the referee that different landscape heterogeneity proxies can describe different landscape characteristics and, indirectly, ecosystem services. Several landscape metrics are available in landscape ecology and Schindler et al., 2008 and 2013, provide a very good description of landscape descriptors useful in ecological studies. We have chosen weighted edge density and the Shannon diversity index as cost-effective strategies for different reasons.

1. The variables selected for this study can be easily observed and understood by landscape planners without requiring complex calculations or analysis. These landscape variables typically relate to fundamental aspects of the landscape, including topography, vegetation, and land use.
2. We tried to briefly describe two important aspects of the landscape, examining mainly aspects related to the landscape composition (e.g., evenness or Shannon diversity index) and configuration (edge density). The landscape metrics used in this study are among the most popular ones used in ecological studies (Tonetti et al., 2023). The Shannon diversity is a good indicator of the balance or unbalance of the amount of each type of land use in the landscape mosaic focused or land cover diversity (potentially associated with the overall biodiversity, due to the number of available habitats) (see, Stein et al. 2014, Kisel et al., 2011). On the other hand, the weighted edge density is strongly associated with the presence of linear elements, such as hedgerows, tree lines, etc, increasing the overall landscape heterogeneity and the border effect. Previous studies have already shown the importance of these two landscape metrics (e.g., Shannon diversity and weighted edge density) as potential biodiversity surrogates (e.g., Morelli et al. 2013). See below the list of studies:

Castro-Caro, J.C., Barrio, I.C., Tortosa, F.S., 2015. Effects of hedges and herbaceous cover on passerine communities in Mediterranean olive groves. Acta Ornithol. 50, 180–192. https://doi.org/10.3161/00016454AO2015.50.2.006

Kisel, Y., McInnes, L., Toomey, N.H., Orme, C.D.L., 2011. How diversification rates and diversity limits combine to create large-scale species-area relationships. Philos. Trans. R. Soc. London B - Biol. Sci. 366, 2514–2525. https://doi.org/10.1098/rstb.2011.0022

Schindler, S., Poirazidis, K., Wrbka, T., 2008. Towards a core set of landscape metrics for biodiversity assessments: A case study from Dadia National Park, Greece. Ecol. Indic. 8, 502–514. https://doi.org/10.1016/j.ecolind.2007.06.001

Schindler, S., von Wehrden, H., Poirazidis, K., Wrbka, T., Kati, V., 2013. Multiscale performance of landscape metrics as indicators of species richness of plants, insects and vertebrates. Ecol. Indic. 31, 41–48. https://doi.org/10.1016/j.ecolind.2012.04.012

Stein, A., Gerstner, K., Kreft, H., 2014. Environmental heterogeneity as a universal driver of species richness across taxa, biomes and spatial scales. Ecol. Lett. 17, 866–880. https://doi.org/10.1111/ele.12277

Morelli, F., 2013. Relative importance of marginal vegetation (shrubs, hedgerows, isolated trees) surrogate of HNV farmland for bird species distribution in Central Italy. Ecol. Eng. 57, 261–266. https://doi.org/10.1016/j.ecoleng.2013.04.043

Morelli, F., Pruscini, F., Santolini, R., Perna, P., Benedetti, Y., Sisti, D., 2013. Landscape heterogeneity metrics as indicators of bird diversity: Determining the optimal spatial scales in different landscapes. Ecol. Indic. 34, 372–379. https://doi.org/10.1016/j.ecolind.2013.05.021

Xu, C., Huang, Z.Y.X., Chi, T., Chen, B.J.W., Zhang, M., Liu, M., 2014. Can local landscape attributes explain species richness patterns at macroecological scales? Glob. Ecol. Biogeogr. 23, 436–445. https://doi.org/10.1111/geb.12108

3. The density/number of patches, shape index, and landscape integrity connectivity may be  correlated with the chosen landscape variables in this study. However, due to the high multicollinearity among predictors, including numerous explanatory variables in the model can lead to overinflated significance. To increase the clarity of our methods, we have now highlighted that:

"From the several landscape metrics available for ecological studies [61,62], we chose two metrics, trying to follow a cost-effective strategy. The two metrics used describe two important landscape characteristics, examining mainly aspects of landscape composition (e.g., evenness or Shannon diversity index) and configuration (weighted edge density). Previous studies have focused on both metrics and shown potential biodiversity surrogacy [63,64]. The Shannon diversity is a good indicator of the land cover diversity. It is potentially associated with the overall biodiversity due to the number of available habitats [65,66]. While the weighted edge density is strongly associated with the presence of linear elements, such as hedgerows, tree lines, etc., increasing the overall landscape heterogeneity, the border effect, and potentially significantly affecting biodiversity [64,67]."

Cited references

Tonetti, V., Pena, J. C., Scarpelli, M. D. A., Sugai, L. S. M., Barros, F. M., Anunciação, P. R., et al. (2023). Landscape heterogeneity: Concepts, quantification, challenges and future perspectives. Environ. Conserv. 127. doi: 10.1017/S0376892923000097

Question 4

4. What does the meaning of '3.2 Figures, Tables and Schemes' in lime 210 on page 5.

Response 4: Thank you very much. We have removed this section and inserted the figures and tables in the appropriate part of the manuscript as per the author's guidelines of the journal.

Question 5

5. Is it 'four' in line 239 on page 7?

Response 5: Thank you very much. We have corrected the number "six" to "four".

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have revised the manuscript carefully, and the manuscript could be published on the current form I think.