the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Jul 2026
Fundamental conditions "indicators" should meet to be objective and to guide practice reliably
Abstract. Over the last 2 decades, a significant body of research has been devoted to finding parameters that could serve as “indicators” of a range of soil properties of interest, including soil health and the many functions/services of soils. Even though the topic had been the object of a sustained debate among ecologists between 2001 and 2010, no one seems to have looked so far into conditions soil indicators should satisfy to be objective, nor is it made clear to users and policy makers which ones are objective and which ones are value-laden and therefore subjective. In the present article, we identify the conditions a soil parameter A needs to satisfy in order to be an objective indicator of another property B. We argue that in addition to A and B being measurable and not merely quantifiable approximately, a testable mechanistic theory needs to exist, which causally relates A and B. In most contexts in which soil indicators are currently used, such a theory appears to be entirely lacking, leading us to call for a significant research effort focusing on theory development.
- Preprint
(1218 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 17 Aug 2026)
- CC1: 'Objectivity is essential in information researchers provide to the public about soil processes (Comment on egusphere-2026-3485)', Philippe Baveye, 11 Jul 2026 reply
-
RC1: 'Comment on egusphere-2026-3485', Moritz Laub, 13 Jul 2026
reply
I just saw that the authors already wrote a public response to some comments that they received by the editor. Hence, I want to express that I formed my opinion before I saw these responses, (and before I saw the comments; even though I would have been able to see them beforehand, I only read the article, not the communication history). So if there are overlaps, (seems to be a bit with comment 1), that is due to the content of the pre-print.
Actual comment:
This suggested forum article sets out to define what would be needed to have meaningful soil property indicators, which is a valuable goal that would certainly merit to be published as a forum article. The article starts interestingly, with the need to have such meaningful and measurable indicators and to define standard measurements to derive these indicators. Also, I find the suggestion to distinguish clearly measurable indicators from value-laden, normative ones a good one.
However, starting with section “2 Conditions of objectivity” and thereafter, I think the article really starts to deviate from the set goal. Instead of clearly defining what would be needed to call an indicator objective and make it useful, the authors criticize what they find subjective and not measurable in the work of others, citing articles that deal with ecosystem services (ESS) that are largely not directly related to soil anymore. This is where the article (in my opinion) turns subjective (addressing mainly the dislike of the authors), and, instead of providing constructive conditions, it criticizes the work of others that is not even directly related to the topic at hand. To me this adds very little value to the debate.
As per the definition of SOIL, Forum articles should stimulate an open debate by presenting new ideas and views of soil as part of the larger Earth system. As such, they must strive to be a “point of departure for future work”.
Hence the focus must be constructive and rather than specify “Don’ts” it should specify the Do`s (e.g. provide clear definitions: what is needed for an objective indicator?, how should it be defined?, give some examples).
So I think if the authors want the article to be considered for publication as a forum article, they should establish clear conditions for suitable soil indicators and provide examples. For example, defining what is meant by “measurable”. Is it sufficient here that it is measurable by a standardized, published, and reproducible method? Or is the practical level of establishment across countries also a factor (e.g., in terms of SOC, Walkley Black may be available across the globe, vs elemental analyzers may not)?
Even the “Don’ts” are rather vague for me – which makes it hard to have a real discussion. It should provide examples of indicators (could be SOC, SOC/clay ratio, microbial diversity or whatever) and highlight why they are good or poor indicators.
One of the core sentences in the conclusion is “We offer a blueprint for how to determine if a given soil indicator can be considered objective.”
This goal is very interesting and a forum article that would this would in fact be very valuable. However, from my perspective, the article does not provide that – it merely criticizes what others (from related fields, not even the same one) have been doing without offering a clear perspective of what should be done. That can hardly be called a blueprint.
So in essence, I think that the goal of the article is very commendable and I would strongly support the publication of such an article as a forum article, if it would achieve that goal. However, in the current form, that is not at all the content of the article (rather it appears to be mostly personal opinion of what people are doing wrong) and so in the current form the article does (in my opinion) not provide the value that is requested to be acceptable as a forum article. Nevertheless, if it would be offer clear definitions, clear examples of measurable vs “pseudo-scientific” methods and some ideas (Do`s) for the way forward, I think it would become a valuable forum article.
Citation: https://doi.org/10.5194/egusphere-2026-3485-RC1 -
AC1: 'Reply to comments of Reviewer 1', Philippe Baveye, 20 Jul 2026
reply
We are very grateful to Moritz Laub for the time he spent analyzing our manuscript, and for his insightful comments about it, especially about ways that we could improve our text. They are very well taken and very much appreciated.
The first of his comments deals with the fact that we cite articles on ecosystem services, which are purportedly “largely not directly related to soil anymore”. Indeed, the example we give in Figure 1, although excerpted from an article dealing specifically with the functions and services of soils, is not directly related to soils. There are many similar figures that we could have taken from the literature on soils, for example from some of the work published recently by Wadoux et al. (2026), to illustrate the fact that maps that are routinely drawn for several ecosystem services of soils are based on very subjective assessments, often based on arbitrarily chosen indicators, and not at all on actual measurements of services or on process-based modelling. It seems that if we pointed this out, this would at least partially answer Moritz’s comment.
One such example would be particularly indicated in that context. It is the estimation of a number of soil services, published last year by Ungaro et al. (2025). Their figure 7 depicts a “wind rose” diagram of model-based estimates of 8 soil-based ecosystem services (SES) associated with different pedolandscape units in the Emilia-Romagna region of Italy. A more detailed figure is provided in a web site put together by the group (at URL: https://ambiente.regione.emilia-romagna.it/en/geologia/soil/soil-planning/soil-ecosistem-services). We reproduce it here in the appended file.
What makes this illustration particularly interesting is that the authors, in the conclusions of their article, candidly discuss a significant shortcoming of the modelling approach they used, based on digital soil mapping (DSM) and pedotransfer functions (PTFs). They point out “the lack of explicit consideration and assessment of how spatial uncertainties propagate through the SES modeling workflow. This means that DSM models and PTFs produce output in terms of potential SES supply indicators maps, but the reliability and confidence of those indicators, given the inherent uncertainties in inputs and model assumptions, are not quantified.” (emphasis added). The imperative need for such quantification was stressed earlier by Hamel et al. (2017) and Baveye (2017), and now appears to be widely accepted in the ecosystem services literature at large. Only the extent to which “validation” of ES assessments actually occurs in practice in the published literature is still an object of debate (Pereira et al., 2025, 2026; Bullock et al., 2026).
The questioning by Ungaro et al. (2025) about the shortcomings of the methodology they used paves the way for us to answer the second comment made by Moritz Laub, namely that our manuscript focuses too much on the “Don’ts” and not enough on the “Do’s”. Not only do Ungaro et al. (2025), very lucidly, point out a shortcoming of their own approach from a perspective that parallels our own, but, even more interestingly, their article was followed by another one, published recently in SOIL by a group of Italian researchers (including Fabrizio Ungaro), in which the authors (Yosef et al., 2026) used a detailed process-based model to investigate “the relationships between soil properties and provision of water regulation ecosystem services across three contrasting pedo-climatic regions in Austria, Italy, and Tunisia”. One of their conclusions is that “several widely used indicators exhibited weaker or inconsistent relationships with water-related processes than commonly assumed”. These included the saturated hydraulic conductivity, often regarded as a key predictor of infiltration rates and water-regulation ecosystem services, and organic matter content as an indicator of water availability. In particular, concerning the latter, they showed that “although higher OM is often expected to enhance water availability and reduce crop water stress, the relationships are not always straightforward”. In addition, their analysis showed that “widely accepted soil property–process relationships are dependent on site-specific soil and climate conditions. They are not universally valid and must be interpreted within the broader context of system complexity, supporting the need for an adaptive, iterative approach to indicator selection”. They point out that “these insights emerge only when […] indicators are developed by explicitly accounting for water-process dynamics and their complex interactions. In this regard, the effectiveness of the analysis may be partly constrained by the lack of robust soil hydraulic property datasets. In view of this, future improvements to the present work will involve conducting the same analysis but focusing exclusively on measured hydraulic properties.”.
To some extent, to respond tongue in cheek to Moritz’s second comment, one could argue that since Yosef et al. (2026) could publish an article in which they criticize currently-used indicators of soil properties and point out what does not seem to work, perhaps the same opportunity should be provided to us as well… In that context, it is not a question of “dislike” on our part to point out that, at this stage, some indicators are subjective. Dislike has nothing to do with it. Farmers and policy-makers, who are looking to soil scientists for answers, need to know whether what we tell them is based on facts and is therefore reasonably secure, or is based on personal opinions. In particular, it is vital for farmers to know to what extent the advice they get is reliable and backed by solid data, so that they can gauge the associated risk. Therefore, we owe it to them to distinguish clearly between what is objective, and what is not, at this stage.
From that perspective, we can use the example provided by Yosef et al. (2026) to clarify what, in the original version of our manuscript, we consider to be “a blueprint for how to determine if a given soil indicator can be considered objective”. Our current description of this blueprint is admittedly brief. For theoreticians, that would be sufficient, but we understand that for anyone else, a little more meat on the bones might be useful. The process-based modelling effort of Yosef et al. (2026) could help us illustrate how that blueprint would operate in practice.
References
Baveye, P. (2017). Quantification of ecosystem services: Beyond all the “guesstimates”, how do we get real data?. Ecosystem Services, 4, 47-49.
Bullock, J., Hooftman, D., Redhead, J., & Willcock, S. (2026). Ecosystem service models are indeed being validated: a response to Pereira et al. Geography and Sustainability, 7(1), 100412.
Hamel, P., & Bryant, B. P. (2017). Uncertainty assessment in ecosystem services analyses: Seven challenges and practical responses. Ecosystem Services, 24, 1-15.
Pereira, P., Inacio, M., Barcelo, D., & Zhao, W. (2025). Ecosystem services mapping and modelling. Where is the validation?. Geography and Sustainability, 6(3), 100286.
Pereira, P., Inacio, M., Barcelo, D., & Zhao, W. (2026). Reply to Bullock et al. (2026) on ecosystem service models are indeed being validated. Geography and Sustainability, 7(2), 100462.
Ungaro, F., Tarocco, P., & Calzolari, C. (2025). Leveraging soil geography for land use planning: Assessing and mapping soil ecosystem services indicators in Emilia-Romagna, NE Italy. Geographies, 5(3), 39.
Wadoux, A. M. C., Debeljak, M., Lagacherie, P., & Creamer, R. E. (2026). Spatial hotspots and bundles of soil functions across Europe. Ecological Indicators, 182, 114481.
Yosef, B. A., Basile, A., Coppola, A., Ungaro, F., Zucca, C., & Bancheri, M. (2026). Soil indicators for ecosystem services: a focus on water regulation. Soil, 12(1), 347-369.
-
AC1: 'Reply to comments of Reviewer 1', Philippe Baveye, 20 Jul 2026
reply
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 60 | 22 | 7 | 89 | 6 | 5 |
- HTML: 60
- PDF: 22
- XML: 7
- Total: 89
- BibTeX: 6
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Objectivity is essential in information researchers provide to the public about soil processes
Philippe C. Baveye
Saint Loup Research Institute, 7 rue des chênes, La Grande Romelière, 79600 Saint Loup Lamairé, France.
Dr. Jasper Wubs, the topic editor assigned to handle the review of our manuscript, has expressed a number of “concerns” in his “Justification” of his decision to post our manuscript as a preprint, and to start the discussion about it. This “Justification” has been so far visible only to the authors and reviewers. Given the nature of the concerns expressed, I think that in the interest of open communication and transparency, they should be accessible to everyone and addressed directly, and some kind of a response should be formulated before reviews are turned in. In the interest of time, and since it is the middle of the Summer and my co-authors are off on holiday with their families, I will provide a personal response below.
As a preliminary remark, I would like to emphasize that our manuscript was submitted as a Forum piece, i.e., as a text meant first and foremost to stimulate a debate on a topic that we think is worth discussing more extensively than is the case at the moment. We recognize fully, from the onset, that some of our colleagues in soil science may disagree with what we are writing and proposing. If we were all in agreement, there would be no need for a Forum piece. In that context, the question that should be asked of the reviewers at this stage is not whether they agree with the conclusions we reach, but rather whether we raise a valid question, and whether our manuscript is formulated in such a way that it is likely to stimulate a constructive debate on this question.
Concern 1: “[it’s] not directly apparent what the intellectual step is that is made here, indicators and subjectivity are central elements in measurement theory, epistemology and philosophy of science - what from that discussion is peculiar (and new) to soil? And what is the step you propose?”
Response: One could debate whether the statement that “indicators and subjectivity are central elements in measurement theory, epistemology and philosophy of science” is accurate. That may be true in some fields. However, in the 45 years since I started reading assiduously on these various topics as they potentially related to research in soil science or education, I have developed a viewpoint on them that does not correspond at all with the above statement.
In any event, that is not really the debate. Our Forum piece focuses on the description of processes occurring in soils, and in that context, the question we ask, about the conditions indicators need to meet to be considered “objective”, has, to our knowledge, not been asked explicitly so far. The closest researchers have come to address it is in the ecology literature between 2000 and 2010. But we have not been able to find any echo in the soil literature of the very interesting debate that took place then, before it eventually faded in the background without resulting in any kind of consensus.
The step we propose to move forward is clearly detailed in our text. It is a renewed emphasis on theory development in soil science to establish clear causal relationships between potential indicators and soil properties.
Concern 2: “objectivity seems to be your guide, how do you relate that against the idea that 'all data are value-laden', an important academic notion in much of the social sciences. It applies directly to the field of soil health/quality/functioning where we can't even agree on the definitions of these central concepts, let [alone] agree on their measurement.”
Response: The proposition that “all data are value-laden” is indeed popular in some disciplines, but one could easily argue that, in the context of soils, it is an exaggeration. Some data, like the 43°C surface temperature my (correctly positioned) outdoors thermometer measured yesterday, are not value-laden. One should not confuse “value-laden” with the perspectival character that is associated with the type of instrument that one chooses to use to measure soil properties. However, although very interesting and worth investigating as well, that is an entirely different, philosophical debate than the one we are trying to stimulate in our manuscript.
Objectivity, for us, is not just a “guide”. Our perspective on objectivity departs to some extent from one that at least some researchers seem to espouse. As, e.g., Hundemer et al. (2026) pointed out recently, there is a perception among various constituencies that researchers, in their pursuit of the advancement of knowledge, should “produce objective science”, but at the same time, they should also “use their science and expertise to support the public good” and in that context, the requirement of objectivity should not be as stringent. Hundemer et al. (2026) refer to “tensions” between those two contrasting facets of research endeavours. Again, that may be true in some disciplines, but I am not convincing that with respect to soil science, the need for objectivity is so different among the complementary activities of researchers.
The question we ask in our Forum piece is, at its root, a very pragmatic one. If we are trying to convince a farmer that s/he needs to increase the microbial biodiversity of her/his soil to increase crop yields or to help mitigate climate change, s/he will immediately ask what evidence we have that there is a clear link between biodiversity and these processes, in other what guarantees we can give her/him that increasing the soil microbial diversity will automatically produce the desired outcome. Even before the quintessential question is discussed of how much that would cost, it is important in practice for the farmer to know whether what we are talking about is a reasonably established fact, and how strong the evidence for it is, or whether it is just a personal opinion. In that context, what we are discussing in our Forum piece is the set of conditions that need to be met for a statement we make to be considered a sound, reliable piece of information. That is needed by practitioners, policy makers, and, I would argue, anyone who deals with soils or benefits from their services, in one way or another. With farmers, the guarantee that there is evidence in support of a new practice does not automatically mean that they will adopt it widely. From my experience, some of them, closely monitored by their neighbours, may try it in a corner of one of their fields, and if the results are conclusive after 2 or more seasons, they may adopt it more widely. But in order to get a foot in the door, there definitely has to be more than a subjective opinion at the onset.
The issue of soil health is a good example in that context. There has been a sustained debate on what the concept means and a number of researchers in recent years have commented on the protracted absence of an operational definition of it (e.g., Stewart et al., 2018; Janzen et al., 2021; Baveye, 2021, 2025; Powlson, 2021; Harris et al., 2022; Matson et al., 2024; Fausak et al., 2024). If, when one tries to convince farmers that they should improve the health of their soils, we are honest with them and admit that, in the soil science community, not only do we not know how to define this concept, but we do not know either how to measure it, the response of the farmers, from my experience, is rarely very enthusiastic. If the government agreed to pay for it, farmers around me would be willing to paint their soils pink with big yellow dots, without asking whether that makes any sense, and the same goes with campaigns to increase “soil health”, but overwhelmingly from the farmers’ perspective, in the absence of sizeable subsidies, full adherence would be difficult to obtain to a concept that is neither defined nor measurable or convincingly profitable. Farm advisors, at least in the part of France where I live, as well as apps targeting farmers in my region, have understood that very well, it seems. Omitting to mention the absence of a consensus in the soil science community about what soil health is, they tend to provide an arbitrary definition of soil health that suits the farmers they interact with, without pointing out the entirely subjective nature of the list of parameters their definition involves…
Concern 3: “Even in something seemingly solid and objective like agricultural soil testing for nutrient concentrations there is huge variation in methods (e.g. type of P extraction), and labs don't always agree in the estimated concentrations, and may get sample rank-orders jumbled up. So, even if such chemistry metrics could be objective indicators in theory, how far is that theory away from daily practice and does that then still qualify under your objectivity paradigm?”
Response: I guess the key word in that concern is the word “seemingly”... If agricultural soil testing for nutrient concentrations were really “solid and objective”, there would not be different methods to assess some of the key parameters, like the available phosphorus concentration, and the answer farmers would get about how much phosphorus they need to add to their soil would not depend on which laboratory they send their samples to, or on the person they consult.
It is difficult for me to imagine that this current state of affairs, as described in this Concern, would be deemed satisfactory to anyone in the soil science community, let alone to those who depend on reliable information, independent of whoever they happen to talk to, to manage their soils. On this particular question, and I would argue on quite a few others about soils, we need to get to a point where we are able to provide reasonably consensual answers. That will clearly require some fundamental research and the development of theory. How long that is going to take is difficult to forecast. Perhaps if soil scientists stopped publishing inordinate amounts of articles promoting endlessly the same sales pitches about soil carbon sequestration, soil health, soil security, etc…, and devoted time instead to designing and carrying out careful, targeted experiments, it would be possible in the not too distant future to reduce significantly the amount of subjectivity in the practical advice we give to farmers and the public at large about soil management.
References
Baveye, P. C. (2021). Soil health at a crossroad. Soil Use and Management, 37(2), 215-219.
Baveye, P. C. (2025). Why a mechanistic theory of soils is crucially important: Another line of supportive argument exists, seldom invoked in soil science. Soil, 11(2), 1131-1140.
Fausak, L. K., Bridson, N., Diaz-Osorio, F., Jassal, R. S., & Lavkulich, L. M. (2024). Soil health–a perspective. Frontiers in Soil Science, 4, 1462428.
Harris, J. A., Evans, D. L., & Mooney, S. J. (2022). A new theory for soil health. European Journal of Soil Science, 73(4), e13292.
Hundemer, S., Loizzo, J., Peterson, A. L., Stoutamire, S., & Paxson, S. B. (2026). Scientific objectivity and the pursuit of the public good. Science and Engineering Ethics. In press
Janzen, H. H., Janzen, D. W., & Gregorich, E. G. (2021). The ‘soil health’metaphor: Illuminating or illusory?. Soil Biology and Biochemistry, 159, 108167.
Matson, A., Fantappiè, M., Campbell, G. A., Miranda-Vélez, J. F., Faber, J. H., Gomes, L. C., ... & Chenu, C. (2024). Four approaches to setting soil health targets and thresholds in agricultural soils. Journal of Environmental Management, 371, 123141.
Powlson, D. S. (2021). Is ‘soil health’meaningful as a scientific concept or as terminology. Soil Use and Management, 37(3), 403-405.
Stewart, R. D., Jian, J., Gyawali, A. J., Thomason, W. E., Badgley, B. D., Reiter, M. S., & Strickland, M. S. (2018). What we talk about when we talk about soil health. Agricultural & Environmental Letters, 3(1), 180033.