Reframing the Debate Surrounding the Yield Gap between Organic and Conventional Farming

Round  1

Reviewer 1 Report

Summary

The paper takes a valuable approach and presents a new framework to analyse the yield gap between organic and conventional agriculture. Given the increasing literature, the overview on the existing studies is helpful and interesting. The authors discuss several drawbacks of the literature and propose a new approach for future analyses.

The paper is an interesting and valuable overview on the debate, which adds important factors to be taken into account. However, in some details, the overview is not systematic enough and misses out some existing approaches to solve the methodological problems, which are indeed existing in the literature.

Major comments

1.     Introduction & Discussion: Generally, the main problem of the political debate on farm types is the implicit assumption, that one could easily upscale a farming system, once having identified that one or the other system is “superior”. However, this “upscaling” assumption (coming e.g. from farmers associations, favouring conventional farming systems and stating that e.g. “Europe should feed the world”…) is not consistent within a market economy, where we seek for individual solutions for an individual problem. The narrative “What can different systems add to the world food situation?” is a more realistic way to tackle the problem. I am not sure, whether it makes sense to take up this point, but from a development policy perspective, it is more important to identify the right system among different systems for specific locations, climates, or farm types etc. My question to the authors would be: To what extent is that consistent with the proposed model?

2.     Table 1:If the variation is important (as stated in the introduction), I would propose to add a column with e.g. a standard error (if presented in the studies). Otherwise, the paper runs into exactly the same trap as described. And some papers are presenting standard errors. Badgley et al. (2006) are presenting standard errors, de Ponti et al. (2012) are presenting ranges, Seufert et al. 2012 are using confidence intervals etc. 

3.     4.1: As the authors state, it is indeed important, to think about additional factors, which could potentially influence (or even distort) comparisons. This is partly described in section 4.1, however reading this section, the text looks a bit unsystematic, especially, since the text is following studies and not according factors. I would recommend, to provide a more systematic review structured according the factors (e.g. climate factors, location, soil, nutrients, varieties, socio-economic factors etc.) identified in the different meta-studies. For an overview on socio-economic effects, see Lakner & Breustedt (2017). 

So far, the text suggests, that there is little awareness on systematic difference, but there is more than presented. To a certain extent, the text is incomplete in that sense, since some of the cited meta studies provide explanatory factors. In some of the meta studies, the authors take into account, which might additionally influence the yield gap beyond the pure system differences. I would recommend to add a bit in section 4.1 and structure it in a similar way, as it is done in section 4.2 and following.

4.     Discussion: One point raised in the discussion is the question, whether we can use an “ecologically sustainable maximum yield” to identify a “true yield gap”. First, it might be hard to identify such a sustainable yield. In crop experiments, one would have to define a “reference systems” (e.g. what is sustainable yield to compare with…). This deems on the one hand methodologically right. However, in practical agriculture working under competitive conditions, this might not feasible. This relates to the previous point 4: If we can identify “comparable conventional systems”; we might get a more realistic picture. 

The proposed new framing is based on one methodological question: Are differences between organic and conventional due to system differences or is part of the yield gap driven by structural factors. The sampling strategy can restrict the interpretation of results, which has been early pointed out by Offermann & Nieberg (2001). So if comparing conventional and organic systems, it is crucial to have a clear description on structural differences (see Offermann & Nieberg 2001, also in Lakner & Breustedt 2017). There are methodological approaches in economics to account for structural differences between organic and conventional farms, such as e.g. matching or selectivity models. The paper still misses a constructive overview. It would be valuable to elaborate, to what extent this is taken into account. Otherwise, the paper is missing a crucial methodological issue.

Minor comments

L23:I would suggest choosing a different wording than “hotly debated”, this sounds a bit odd. 

I would propose to add the paper of Kniss et al. (2016) to the literature overview. 

Literature:

Kniss, A.R., S.D. Savage & R. Jabbour (2016): Commercial Crop Yields Reveal Strengths and Weaknesses for Organic Agriculture in the United States. PLoS ONE 11(8): e0161673. doi: 10.1371/journal.pone.0161673

Lakner S. & G. Breustedt (2017):  Efficiency Analysis of Organic Farming Systems –A Review of Concepts, Topics, Results and Conclusions, German Journal of Agricultural Economics 66, 2: 85-108

Offermann, F. & H. Nieberg (2001): Economic performance of organic farms in selected European countries: situation, development and determinants (in German). In: German Journal of Agricultural Economics (Agrarwirtschaft) 50 (7): 421-427.

Author Response

Response to Reviewer 1

Major comments

1.     Introduction & Discussion: Generally, the main problem of the political debate on farm types is the implicit assumption, that one could easily upscale a farming system, once having identified that one or the other system is “superior”. However, this “upscaling” assumption (coming e.g. from farmers associations, favouring conventional farming systems and stating that e.g. “Europe should feed the world”…) is not consistent within a market economy, where we seek for individual solutions for an individual problem. The narrative “What can different systems add to the world food situation?” is a more realistic way to tackle the problem. I am not sure, whether it makes sense to take up this point, but from a development policy perspective, it is more important to identify the right system among different systems for specific locations, climates, or farm types etc. My question to the authors would be: To what extent is that consistent with the proposed model?

Thank you for your feedback! While these are excellent points, the scope of this paper is not to identify an ideal agricultural system for any given set of geographic/climatic/socioeconomic conditions, but rather to reframe the common idea of yield gaps in particular. We very much agree with your observation that individualized solutions are critically necessary. Within the framework of this paper, our proposal of a varying threshold for ecologically sustainable maximum yields is very much consistent with the points you make above: starting conditions in terms of natural resources and climate, as well as the demands and values of local producers and consumers, vary dramatically in agricultural systems worldwide. Thus, the ecologically sustainable yield threshold (and consequently the size of the true yield gap) varies, as we have tried to show in Figure 2.

2.     Table 1: If the variation is important (as stated in the introduction), I would propose to add a column with e.g. a standard error (if presented in the studies). Otherwise, the paper runs into exactly the same trap as described. And some papers are presenting standard errors. Badgley et al. (2006) are presenting standard errors, de Ponti et al. (2012) are presenting ranges, Seufert et al. 2012 are using confidence intervals etc.

Thank you for the suggestion. We did not include measures of variance for individual studies (as you point out above), because our intention is not to address individual studies but rather to describe meta-analyses by category of crop. Our observation about variation between studies was meant to point out the importance of additional factors in affecting the yield gap, and adding metrics for the variance between single observations within studies included in the meta-analyses would not strengthen this point.

3.     4.1: As the authors state, it is indeed important, to think about additional factors, which could potentially influence (or even distort) comparisons. This is partly described in section 4.1, however reading this section, the text looks a bit unsystematic, especially, since the text is following studies and not according factors. I would recommend, to provide a more systematic review structured according the factors (e.g. climate factors, location, soil, nutrients, varieties, socio-economic factors etc.) identified in the different meta-studies. For an overview on socio-economic effects, see Lakner & Breustedt (2017). 

So far, the text suggests, that there is little awareness on systematic difference, but there is more than presented. To a certain extent, the text is incomplete in that sense, since some of the cited meta studies provide explanatory factors. In some of the meta studies, the authors take into account, which might additionally influence the yield gap beyond the pure system differences. I would recommend to add a bit in section 4.1 and structure it in a similar way, as it is done in section 4.2 and following.

Thank you for the suggestion. While additional factors are quite important, as we note, we intentionally structured section 4.1 by the date of publication of the larger studies in order to show the evolution of the field from early models to more contemporary approaches. We agree that a review of yield comparisons structured by additional factors would be useful, and socio-economic effects such as those addressed by Lakner and Breustedt would be a valuable component of such a review, but that is outside the intended scope of this section.

We appreciate the suggestion to acknowledge that previous meta-analyses have indeed included explanatory variables in some cases and agree that this will improve the manuscript. We have added sentences to this effect in that section.

4.     Discussion: One point raised in the discussion is the question, whether we can use an “ecologically sustainable maximum yield” to identify a “true yield gap”. First, it might be hard to identify such a sustainable yield. In crop experiments, one would have to define a “reference systems” (e.g. what is sustainable yield to compare with…). This deems on the one hand methodologically right. However, in practical agriculture working under competitive conditions, this might not feasible. This relates to the previous point 4: If we can identify “comparable conventional systems”; we might get a more realistic picture. 

The proposed new framing is based on one methodological question: Are differences between organic and conventional due to system differences or is part of the yield gap driven by structural factors. The sampling strategy can restrict the interpretation of results, which has been early pointed out by Offermann & Nieberg (2001). So if comparing conventional and organic systems, it is crucial to have a clear description on structural differences (see Offermann & Nieberg 2001, also in Lakner & Breustedt 2017). There are methodological approaches in economics to account for structural differences between organic and conventional farms, such as e.g. matching or selectivity models. The paper still misses a constructive overview. It would be valuable to elaborate, to what extent this is taken into account. Otherwise, the paper is missing a crucial methodological issue.

Thank you for your constructive suggestions! Our intention with this piece was not to choose an appropriate matched control for each organic farm or farming system, but to fundamentally rethink the concept of comparing conventional and organic systems. Whether yield gaps in the existing literature are due to structural or systemic differences between organic and conventional farms is thus a question we leave to experts in another field; we appreciate your recommendation to refer to Offermann and Nieberg (2001) and Lakner and Breustedt (2017) and agree that the economic perspective is a valuable complement to the views we present in this paper. However, methodological issues of past studies (particularly related to agricultural economics, which is not our area of expertise) are not the focus of this paper; rather, we seek to move towards a new paradigm in which comparisons are made only between systems premised upon the same values.

Minor comments

L23:I would suggest choosing a different wording than “hotly debated”, this sounds a bit odd.

Thank you for the suggestion! We have revised the wording.

I would propose to add the paper of Kniss et al. (2016) to the literature overview.

Thank you for the suggestion! We have added this paper to our literature review (lines 95-101).

Reviewer 2 Report

This paper looks at the issue of whether organic can feed the world. Instead of simply reviewing the evidence, the authors ask whether this is the right question – a much more important and interesting issue. I find the arguments well developed in a well written, thought-provoking paper. The overall problem is that the comparison is between organic and worst case, polluting, non-organic farming, and so I see the contrasts as being much more subtle than implied by this paper. My detailed comments should be seen as helpful suggestions.

The title is clear, but rethink the key words as they are vague and some repeat the title. The abstract is correct, but underplays the amount of quantitative analysis. Perhaps the introduction should also cover the nature and limitations of the contrasts used in this paper.

Section 2 – I would find it helpful in 2.1 to clarify whether the yield gaps are in a particular field in a particular year, or averaged across a rotation. The issue of rotations and agribiodiversity and integrating arable and livestock are not really covered

Section 3 – Another point here is that the diversity of non-organic farming is very high, and if anything the gap between organic and non-organic practices is closing. Many farmers would not recognise the vision of conventional farming presented here: I am aware of non-organic farmers who farm according to paradigms you ascribe to organic farmers. You write “A wheat field planted from fencepost to fencepost, turbocharged 115 with synthetic nitrogen, and kept weed- and almost insect-free with toxic chemicals may produce 116 high grain yields, but at what cost?” this statement does not represent all farms by any means.

Section 4 – Lovely to see reference to de Wit! The model (Fig 1) again oversimplifies. For example, many organic farms use tractors which require fossil-based inputs. Again, some non-organic farmers are increasing soil carbon, see also climate-smart agriculture. 4.3 l 211 – This statement is clouded by the rotaitons issue, which does affect transformation of natural resources.

Discussion – There is much discussion on benchmarks for agricultural performance that are not just about yield – see recent literature on measuring the sustainable intensification of agriculture. The second poin about raising poor performance applies to farming in general. Breeding for rhizosphere rtaits is not just related to organic systems.

Author Response

Response to Reviewer 2

This paper looks at the issue of whether organic can feed the world. Instead of simply reviewing the evidence, the authors ask whether this is the right question – a much more important and interesting issue. I find the arguments well developed in a well written, thought-provoking paper.

Thank you, much appreciated!

The overall problem is that the comparison is between organic and worst case, polluting, non-organic farming, and so I see the contrasts as being much more subtle than implied by this paper. My detailed comments should be seen as helpful suggestions.

We agree that the contrasts were somewhat overstated in the first draft, and the comparison was not meant to be that polarizing. We are willingly considering your comments and working to revise overstated language throughout the manuscript.

The title is clear, but rethink the key words as they are vague and some repeat the title.

Thank you for the suggestion! We have revised the key words to avoid redundancy and include additional key phrases.

The abstract is correct, but underplays the amount of quantitative analysis. Perhaps the introduction should also cover the nature and limitations of the contrasts used in this paper.

Thank you for this feedback!  We changed the wording and made clear that those quantitative analyses are of great value.

Section 2 – I would find it helpful in 2.1 to clarify whether the yield gaps are in a particular field in a particular year, or averaged across a rotation.

Good point. We have added wording (lines 55-56 and Table 1 caption) to clarify that in these meta-analytical approaches, yield gaps are calculated across all plots and years of the respective study.

The issue of rotations and agribiodiversity and integrating arable and livestock are not really covered

This is an excellent observation. We have added a concluding paragraph to section 2 (lines 102-109) to address these issues and why it is difficult to include them in a meta-analytical framework.

Section 3 – Another point here is that the diversity of non-organic farming is very high, and if anything the gap between organic and non-organic practices is closing. Many farmers would not recognise the vision of conventional farming presented here: I am aware of non-organic farmers who farm according to paradigms you ascribe to organic farmers. You write “A wheat field planted from fencepost to fencepost, turbocharged 115 with synthetic nitrogen, and kept weed- and almost insect-free with toxic chemicals may produce 116 high grain yields, but at what cost?” this statement does not represent all farms by any means.

Thank you for these observations! You are right that the conventional-organic comparison was presented as a dichotomy rather than a continuum, and differences exacerbated to make a point. We have revised this section to tone down language where necessary and added a sentence to make it clear that there is increasing overlap between the systems.

Section 4 – Lovely to see reference to de Wit!

The model (Fig 1) again oversimplifies. For example, many organic farms use tractors which require fossil-based inputs. Again, some non-organic farmers are increasing soil carbon, see also climate-smart agriculture. 4.3 l 211 – This statement is clouded by the rotaitons issue, which does affect transformation of natural resources.

Thank you for these observations. We have added wording (lines 238-242) to highlight that beneficial practices (rotations, climate-smart agriculture) are often implemented in non-organic systems. Furthermore, we changed the figure and made clear that it is a simplification.

Discussion – There is much discussion on benchmarks for agricultural performance that are not just about yield – see recent literature on measuring the sustainable intensification of agriculture.

Thank you – we feel that including these strengthens the manuscript! We have expanded the discussion to include four references that have developed different approaches to measure sustainable intensification and the balance between productivity and environmental impacts.

The second point about raising poor performance applies to farming in general. Breeding for rhizosphere traits is not just related to organic systems.

Excellent observations! We have revised the paragraph about raising minimum yields (lines 387-397) so that the language is more inclusive. And we agree wholeheartedly about the need to breed for rhizosphere traits in all systems; lines 417-419 have been added to underscore this point.

Reviewer 3 Report

The paper provides a comprehensive review of the literature on yield comparison trials in conventional and organic production systems and delves into the numerous factors that create difficulty in making unbiased comparisons. The authors argue that reframing the question of whether organic can feed the world is premature and that other questions (such as ecologically sustainable yield vs. maximum yield) should be regarded as more pertinent to the debate. The authors also present a new model to support the reframing of the question.

The literature review seems comprehensive and there is an effort to tease apart differences among crops as well as whether production takes place in developed vs developing countries. They argue that research efforts to close the yield gap be focused on developing countries where the greatest gains can be made.

On lines 112-114, the authors argue that conventional focuses on stakeholder “demands” whereas organic focuses on “needs”. It is unclear what definition of demand is used here. In an economic sense, demand refers to consumers’ willingness to pay, while in a conventional sense, it means an insistent request. The statement that “Organic agriculture seeks to meet [stakeholder] needs…” comes from philosophical tenets of organic agriculture. “Needs” as used here implies a basic right. To me, this is an “apples” and oranges” comparison and the sentence needs to be restated in a direct comparison (if that is the author’s intent – for example, both organic and conventional can be described in similar economic terms), or the authors need to go into greater detail on what they actually mean. I think that perhaps what authors are getting at is that organic agriculture is based on a set of philosophical statements while there is no such corresponding basis for any form of conventional agriculture.

Lines 385-388: The authors argue that breeding for rhizosphere traits is a more ecologically sustainable process. One or more references are needed here as I am not aware of much work yet supporting this claim.

Author Response

Response to Reviewer 3

The paper provides a comprehensive review of the literature on yield comparison trials in conventional and organic production systems and delves into the numerous factors that create difficulty in making unbiased comparisons. The authors argue that reframing the question of whether organic can feed the world is premature and that other questions (such as ecologically sustainable yield vs. maximum yield) should be regarded as more pertinent to the debate. The authors also present a new model to support the reframing of the question.

Thank you, much appreciated!

The literature review seems comprehensive and there is an effort to tease apart differences among crops as well as whether production takes place in developed vs developing countries. They argue that research efforts to close the yield gap be focused on developing countries where the greatest gains can be made.

Agreed! It is not only meant to be raised in the developing world but need to be raised in the developed world as well. We changed wording in order to make it clearer.

On lines 112-114, the authors argue that conventional focuses on stakeholder “demands” whereas organic focuses on “needs”. It is unclear what definition of demand is used here. In an economic sense, demand refers to consumers’ willingness to pay, while in a conventional sense, it means an insistent request. The statement that “Organic agriculture seeks to meet [stakeholder] needs…” comes from philosophical tenets of organic agriculture. “Needs” as used here implies a basic right.

To me, this is an “apples” and oranges” comparison and the sentence needs to be restated in a direct comparison (if that is the author’s intent – for example, both organic and conventional can be described in similar economic terms), or the authors need to go into greater detail on what they actually mean.

I think that perhaps what authors are getting at is that organic agriculture is based on a set of philosophical statements while there is no such corresponding basis for any form of conventional agriculture.

Agreed! We changed the wording to make clear what is meant here. Thank you.

Lines 385-388: The authors argue that breeding for rhizosphere traits is a more ecologically sustainable process. One or more references are needed here as I am not aware of much work yet supporting this claim.

Thank you for pointing out the need for references. We have added four references to support this section (lines 405-410).

Round  2

Reviewer 1 Report

Especially points 3 in my last review is critical and definitely needs consideration and improvement. The text should be reconsidered in that sense, otherwise, it would give a wrong reflection on the cited meta-analyses. 

In order to get it more clear, the article states, 

“The meta-analytical design, however, falsely assumes that conventional-organic comparisons 32 can be reduced to a single number and ignores the underlying causes of yield differences.”

This is actually not the assumption or purpose of meta-analyses and this is not like I would read the cited literature. From my perspective, the purpose of meta studies is to show the broad variation within the literature given a number of established statistical key figures. Most studies provide mean values, yes, however, they also show standard deviations (which are ignored in the article), and they do some more statistical analyses. 

The main argument in the paper is, that it is necessary to also include ecosystem services ESS (or as a “very important ESS” soil fertility?). However, looking into some of the cited paper (e.g. Badgley et al. 2007, Seufert et al. 2012, Ponisio et al. 2014, Kniss et al. 2016), their research design is actually considering ESS or other factors, which could partly explain yield gaps. 

So in that sense, the paper is oversimplifying the literature on meta-analyses and this is why I was asking to document, which factors are taken into account by meta analyses.

For instance, Ponisio et al. (2014) states:

„Given that there is such a diversity of management practices used in both organic and conventional farming, a broad-scale comparison of organic and conventional production may not provide the most useful insights for improving management of organic systems.[…] These yield differences dropped to 9+4% and 8+5% when diversification techniques (multi-cropping and crop rotations, respectively) were used.” (Ponisio et al. 2014)

Isn’t this exactly the framing, that the authors suggest? Some other examples can be found as well in the three other mentioned studies. If providing a new “framing” for yield gaps, the article should still give a realistic and fair view on meta studies. 

The provided view on meta-studies is largely oversimplifying the purpose of meta studies. None of the cited articles are only focused on these key figures, they discuss or even explicitly analyze ESS or other determining factors like e.g. location, soil properties or even socio-economic factors.

Obviously, the problem is, that meta-studies can be misunderstood in that sense. Meta analysis is a method, and (as far as I can judge), most of the meta studies are correct (apart from the critic on the paper on Badgley et al....). The paper should be fair to the scientific contribution of other authors, the problem is rather the interpretation of meta-studies in the political arena.

Therefore, the method of meta analysis should be presented absolutely correct, also in other parts of the article, otherwise, the article remains suggestive in this point and would rather be a "political paper", which is not suited for a scientific journal. So, please check, whether the text is 100% correct and fair in on the method of meta-analyses in all parts of the paper.

Figure 1: Adding to my previous points, I would strongly recommend returning to version 1 of this figure. To my best knowledge, soil fertility is regarded as an ecosystem service. So, the figure seems to certain extent arbitrary – why soil fertility and not another ESS? Maybe, this is something to be discussed with the editor, but I would think, that the figure should be general and not add soil fertility as a specific case of another ESS.

Author Response

Dear Editor and Reviewers,

Thank you for your insightful suggestions for improving our manuscript. We have taken your comments into consideration, particularly focusing on our treatment of meta-analytical studies, and present a revised version of the manuscript for review. Below, please find our responses to specific issues (reviewer comments in italics).

Especially points 3 in my last review is critical and definitely needs consideration and improvement. The text should be reconsidered in that sense, otherwise, it would give a wrong reflection on the cited meta-analyses. 

In order to get it more clear, the article states, 

“The meta-analytical design, however, falsely assumes that conventional-organic comparisons 32 can be reduced to a single number and ignores the underlying causes of yield differences.”

This is actually not the assumption or purpose of meta-analyses and this is not like I would read the cited literature. From my perspective, the purpose of meta studies is to show the broad variation within the literature given a number of established statistical key figures. Most studies provide mean values, yes, however, they also show standard deviations (which are ignored in the article), and they do some more statistical analyses. 

The main argument in the paper is, that it is necessary to also include ecosystem services ESS (or as a “very important ESS” soil fertility?). However, looking into some of the cited paper (e.g. Badgley et al. 2007, Seufert et al. 2012, Ponisio et al. 2014, Kniss et al. 2016), their research design is actually considering ESS or other factors, which could partly explain yield gaps. 

So in that sense, the paper is oversimplifying the literature on meta-analyses and this is why I was asking to document, which factors are taken into account by meta analyses.

For instance, Ponisio et al. (2014) states:

„Given that there is such a diversity of management practices used in both organic and conventional farming, a broad-scale comparison of organic and conventional production may not provide the most useful insights for improving management of organic systems.[…] These yield differences dropped to 9+4% and 8+5% when diversification techniques (multi-cropping and crop rotations, respectively) were used.” (Ponisio et al. 2014)

Isn’t this exactly the framing, that the authors suggest? Some other examples can be found as well in the three other mentioned studies. If providing a new “framing” for yield gaps, the article should still give a realistic and fair view on meta studies. 

The provided view on meta-studies is largely oversimplifying the purpose of meta studies. None of the cited articles are only focused on these key figures, they discuss or even explicitly analyze ESS or other determining factors like e.g. location, soil properties or even socio-economic factors.

Obviously, the problem is, that meta-studies can be misunderstood in that sense. Meta analysis is a method, and (as far as I can judge), most of the meta studies are correct (apart from the critic on the paper on Badgley et al....). The paper should be fair to the scientific contribution of other authors, the problem is rather the interpretation of meta-studies in the political arena.

Therefore, the method of meta analysis should be presented absolutely correct, also in other parts of the article, otherwise, the article remains suggestive in this point and would rather be a "political paper", which is not suited for a scientific journal. So, please check, whether the text is 100% correct and fair in on the method of meta-analyses in all parts of the paper.

Upon revisiting these sections, we agree that the previous treatment of meta-analytical studies was not as nuanced as it should have been. Meta-analytical studies do indeed allow the consideration of moderating variables and a range of values, even if the public perception of their findings is not always so nuanced. We have clarified that the meta-analytical method is indeed highly useful at summarizing a number of studies into average yield gaps and ranges, including how those ranges vary in response to moderating variables such as fertilization rates and rotational diversity. We feel that these changes (both in the Introduction and the section on meta-analyses) have addressed the criticism of unfair treatment of meta-analyses and showed that these studies are quite valuable.

Our objective with this paper is not to conduct a meta-analysis ourselves, however, which is why we did not provide a table summarizing the moderating variables considered in each study and how they contribute to yield gaps. Instead, we discuss meta-analyses in order to 1) highlight their important contribution to previous organic-conventional debates, and 2) call into question the idea that conventional agriculture can readily be used as an appropriate benchmark for organic agriculture. We hope that our revisions have made this more clear in the text.

Figure 1: Adding to my previous points, I would strongly recommend returning to version 1 of this figure. To my best knowledge, soil fertility is regarded as an ecosystem service. So, the figure seems to certain extent arbitrary – why soil fertility and not another ESS? Maybe, this is something to be discussed with the editor, but I would think, that the figure should be general and not add soil fertility as a specific case of another ESS.

Thank you for your recommendation. We have returned to a slightly revised version of Figure 1 as suggested.

Thank you for your consideration of our revised manuscript!

Sincerely,

Klaus-Peter Wilbois & Jennifer Schmidt