the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 01 Apr 2025
What if publication bias is the rule and net carbon loss from priming the exception?
Abstract. Priming effects in soil science describe the influence of labile carbon inputs on rates of microbial soil organic matter mineralisation, which can either increase (positive priming) or decrease (negative priming). While both positive and negative priming effects occur in natural ecosystems, the latter is less documented in the peer-reviewed literature and the overall impact of priming effects on the carbon balance of vegetated ecosystems remains elusive. Here, we highlight three aspects which need to be discussed to ensure (rhizosphere) priming effects are correctly perceived in their ecological context and measured at appropriate scales: (i) We emphasize the importance of evaluating net C balances because usually experimental C inputs exceed C losses meaning even positive priming doesn’t cause net C-loss; (ii) We caution against publication bias, which forces overrepresentation of positive priming effects, neglects negative or no priming, and potentially misguides conclusions about C loss; and (iii) We highlight the need to distinguish between general priming effects and rhizosphere- specific priming, which differ in their scale and driving factors, and hence require different methodological approaches. Future research should explore potential discrepancies between laboratory and field studies and examine the role of rhizosphere priming in nutrient cycling and plant nutrition.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(612 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(612 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1067', Anonymous Referee #1, 22 Apr 2025
-
AC1: 'Reply on RC1', Jenny Michel, 12 May 2025
Reply to RC1: 'Comment on egusphere-2025-1067', Anonymous Referee #1
The paper submitted by Michel et al. explores priming effects in soils, which refer to the influence of labile carbon inputs on the mineralization of organic matter by microbes. These effects can be positive (increased mineralization) or negative (reduced mineralization). The article highlights three key aspects to better understand these effects in their ecological context: (i) Evaluation of net carbon balances (ii) Publication bias and (iii) Difference between general priming effects and rhizosphere-specific priming. The article calls for a more nuanced approach to priming research, encouraging the publication of studies on negative or neutral effects, carefully evaluating publication biases, and distinguishing general priming effects from rhizosphere-specific effects. It also emphasizes the importance of conducting field studies to better understand these phenomena under natural conditions.
I think the points developed by the authors are fair and worth to be published but the current version of the paper is a bit disappointing since it skims over the issues without going into depth. It may come from the format which is not very clear. Is it a review or is it an opinion paper? If this is a review, much more literature must be cited if this an opinion paper, the ideas developed must be more attractive and be inspiring for the community to change their ways of doing priming research.
>> We thank the reviewer for their time and feedback on this manuscript. We apologies if the format is confusing, it is a “forum article” (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. Purely speculative contributions are discouraged. Manuscript composition is free of parameters, although it is requested that forum articles should be short (less than 2500 words, with a maximum of three figures and tab items-> https://www.soil-journal.net/about/manuscript_types.html). The grains of thought we wish to sow with this article are (i) the increasing evidence that not even positive priming causes net C loss, (ii) the possibility that direction and magnitude of priming effects could be misrepresented due to publication bias and (iii) to scale priming to ecosystem processes methodological changes are needed. The first two points need to be discussed more in the community because the (possibly false) perception of positive priming being a “rule” causes (esp. early career) researchers to repeat and/or discard experiments where they observe negative priming because they think the experiment “didn’t work” or they “made a mistake”. This is a dire situation, and we hope anyone reading the article and having some negative priming data in the corner of their desktop take the time to write it up for peer-reviewed publication because negative priming is a result just as valuable as positive priming. Moreover, if the C-impact of priming is negligible, this means that we can safely consider priming in relation to plant nutrition, which offers many exciting opportunities for future research especially in the field of agronomy. The inspiration the attentive reader could draw from this forum article is therewith two-fold: i) Carefully evaluate their own chain-of-thought and those in the papers they read: Is a net carbon balance provided (this is imperative to talk about net C loss or gain)? Are the hypothesis allowing the option of negative priming occurring? Does the spatial and/or temporal extent over which data is collected match the conclusions / generalizations drawn based on data collected from the study? ii) Carefully think about how future experiments could start with unbiased hypothesis and ideally could include living plants (we'll make some specific suggestions for this in the revised manuscript).
For instance, the authors wrote that “there is little empirical evidence for net C losses”, I fully agree with the statement but it should be better explained why priming is sometimes see as a mechanism leading to net losses and show some papers that suggest it and explain why they are mistaken. So far the authors mostly cite papers that do not observe any net loss.
>> Thanks, we’ll clarify this point in the revised manuscript.
The second section is simply a re-analysis of the data from Xu et al, which merely confirms the conclusions of the original paper. In my opinion, this does not add much to the message of Xu et al. More details on the effects of bias should be provided and some ideas to avoid them should be also proposed.
>> Thanks, we’ll clarify in the manuscript why we think that a meta-meta-analysis with a global average estimate of 10.7% PE (estimated effect size (log-transformed response ratio) of 0.1022 (CI95: 0.0740, 0.1305) is too close to zero priming to state a global rule about the direction of priming given that none of the underlying meta-analysis has been corrected for publication bias. Our re-analysis dropped the author’s priming estimate of +37% to +10.7%, which is 2/3 less than the original estimate (an over-estimation of 60%). It is a huge difference! If we could do a similar exercise of bias correction on the 12 original meta-analysis, it seems possible to very likely that the result would be a global negative to no priming estimate. We would have loved to do this analysis, but unfortunately not all data from the primary meta-analysis was available upon reasonable request. (We were particularly curious about the meta-analysis by Li et al. (2023) which reports exclusively negative priming in response to temperature change but appears with all positive effect sizes in Xu et al.. We would have also loved to better understand how the meta-meta-analysis accounted for the re-appearance of certain primary research papers across the different meta-analysis. For example, Luo et al. 2016 contains n=37 primary studies and n=30 of them re-appear in Sun et al., 2019 and n=43 of the Sun studies re-appear in Yan et al., 2023. The supplementary material of Xu’s meta-meta-analysis includes a redundancy matrix which gives the number of overlap of primary studies included between meta-analysis, but it is not entirely clear how this has been accounted for in the effect sizes (percent change) provided.) We’ll happily dive a bit more into biases of the priming debate in the revised manuscript.
Then the third section is mainly focused on the importance of using intact soils instead of sieved soils whereas the title of the section suggests that RPE and PE should be treated differently. So far this section is not really convincing mostly because it lacks more clear examples and it needs some suggestions to improve the current methods.
>> Thanks, we’ll clarify these points in the revised manuscript. The main point is that many rhizosphere priming studies are not even including a rhizosphere. Sieving soils indeed may also be a problem, not only because it fundamentally changes water hydraulics and therewith plant nutrient uptake (and therewith priming), but also because it possibly affects fungi more than bacteria as the usually used 2mm sieve destroys fungal hyphae but keeps bacteria cells intact. We have some suggestions to improve the current methods, but the best ideas could probably come from an open debate, which is the primary goal of this forum article.
Citation: https://doi.org/10.5194/egusphere-2025-1067-AC1
-
AC1: 'Reply on RC1', Jenny Michel, 12 May 2025
-
RC2: 'Comment on egusphere-2025-1067', Anonymous Referee #2, 24 Apr 2025
In their manuscript with a promising title “What if publication bias is the rule and net carbon loss from priming the exception?” the authors aim to address an important aspect of the relevance of priming. The main point is that priming (i.e. enhanced or reduced mineralization of native soil organic carbon) occurs as an elemental process but the net carbon balance is in most cases not considers in the literature. I agree that this causes a biased understanding of the priming in relation to total carbon changes of soils. In fact, I agree that priming, especially positive, is not on relevant magnitudes to off-set the actual input that is required to initiate priming in the first place. I also generally agree that our understanding is biased due to studies that used “artificial” organic matter inputs and that we have a limited understanding of the complex effects of the whole rhizosphere effect and associated organic matter inputs but also biogeochemical shifts. Overall, I agree that such statements need to be made clearly in order to facilitate future research focuses. However, I have to be honest that I was a little disappointed that the authors remain on a very broad level with their discussion. In addition, it is not clear if this is supposed to be a review or opinion and I miss any clear statement how we should move on.
The first chapter summarizes very briefly that positive priming is not affecting the net carbon balance.
The second chapter re-analysis the study presented by Qin et al (2024) and Xu et al. (2024). In fact, the authors repeat the statement by Qin et al (2024) who stated that the net balance is mainly positive. Therefor, this is supporting the authors point but not providing new information that is shown in Figure 1a that only shows the included studies in Qin et al. Regarding the Xu et al. (2024) study, the chapter in the manuscript is quite technical how the authors re assessed the data and what funnel plots are (line 93-123, which is a lot regarding the short manuscript). The authors provide an improved interpretation of the same plot that is discussed by Xu et al in the supplement figure 2. The authors apply the standard error on ln transformed the priming effect observed in the studies. Technically this seems correct. However, this seems to be rather a response to the Xu et al paper than making a new point. In fact, the Xu et al in the supplement figure 2 already shows the asymmetry that most studies are reporting positive priming effects, when I visually assess the deviation from x=0 in the plot (see plot below). The authors finally estimate that only six studies with moderate or negative priming effects are need to overcome the publication bias. They state in line 137 “It would be interesting to recalculate a global PE estimate from the primary research data of all underlying meta-analysis corrected for publication bias.”. This raises the question why the authors are not providing this. Finally, the authors just encourage to publish negative and moderate priming effects.
The third chapter claims that rhizosphere priming and general priming are different. Yes, it is important that we include more complexity and understand the rhizosphere effects better. It is a major challenge to study such effects in intact systems that include the whole rhizosphere. This is already discussed in other studies and should be clear. The authors just state that we need to understand rhizosphere priming in natural ecosystems, but they do not provide any way forward to do so. Therefore, I do not see any substantial addition to the discussion based on the very broad and short chapter here.
In addition, the authors do not provide any clear definition of the priming effect in the first place. For example, in line 15 it is stated: “Priming effects in soil science describe the influence of labile carbon inputs on rates of microbial soil organic matter mineralisation, which can either increase (positive priming) or decrease (negative priming).” Therefore, they miss the important aspect that priming is affecting the native or present soil organic matter after the addition of new organic matter.
Funnel plot from Xu et al (2024, supplement) with added x=0 vertical line
Citation: https://doi.org/10.5194/egusphere-2025-1067-RC2 -
AC2: 'Reply on RC2', Jenny Michel, 12 May 2025
Reply to RC2: 'Comment on egusphere-2025-1067', Anonymous Referee #2
In their manuscript with a promising title “What if publication bias is the rule and net carbon loss from priming the exception?” the authors aim to address an important aspect of the relevance of priming. The main point is that priming (i.e. enhanced or reduced mineralization of native soil organic carbon) occurs as an elemental process but the net carbon balance is in most cases not considers in the literature. I agree that this causes a biased understanding of the priming in relation to total carbon changes of soils. In fact, I agree that priming, especially positive, is not on relevant magnitudes to off-set the actual input that is required to initiate priming in the first place. I also generally agree that our understanding is biased due to studies that used “artificial” organic matter inputs and that we have a limited understanding of the complex effects of the whole rhizosphere effect and associated organic matter inputs but also biogeochemical shifts. Overall, I agree that such statements need to be made clearly in order to facilitate future research focuses.
>> We thank the reviewer for their time and feedback on this manuscript. We apologise for the brevity and confusion about the “forum” format and will be more specific about possible avenues of future research (which requires that we critically evaluate the conclusions of past and present research which is neglecting negative priming).
However, I have to be honest that I was a little disappointed that the authors remain on a very broad level with their discussion. In addition, it is not clear if this is supposed to be a review or opinion and I miss any clear statement how we should move on.
>> Thanks, reviewer 1 said the same thing.
The first chapter summarizes very briefly that positive priming is not affecting the net carbon balance.
>> Yes.
The second chapter re-analysis the study presented by Qin et al (2024) and Xu et al. (2024). In fact, the authors repeat the statement by Qin et al (2024) who stated that the net balance is mainly positive. Therefor, this is supporting the authors point but not providing new information that is shown in Figure 1a that only shows the included studies in Qin et al. Regarding the Xu et al. (2024) study, the chapter in the manuscript is quite technical how the authors re assessed the data and what funnel plots are (line 93-123, which is a lot regarding the short manuscript). The authors provide an improved interpretation of the same plot that is discussed by Xu et al in the supplement figure 2. The authors apply the standard error on ln transformed the priming effect observed in the studies. Technically this seems correct. However, this seems to be rather a response to the Xu et al paper than making a new point. In fact, the Xu et al in the supplement figure 2 already shows the asymmetry that most studies are reporting positive priming effects, when I visually assess the deviation from x=0 in the plot (see plot below). The authors finally estimate that only six studies with moderate or negative priming effects are need to overcome the publication bias. They state in line 137 “It would be interesting to recalculate a global PE estimate from the primary research data of all underlying meta-analysis corrected for publication bias.”. This raises the question why the authors are not providing this. Finally, the authors just encourage to publish negative and moderate priming effects.
>> We thank the reviewer for acknowledging the improvements our re-analysis brings to the interpretation. We would have been delighted to re-analyse the underlying meta-analysis, including the C-in vs. out-puts, study duplication, publication bias etc. but the data was not available for all articles. We are not sure if it is realised in the community how strong the impact of confirmation and conformity bias is? The publication of negative and no priming effects needs to be encouraged strongly because some researchers do not publish negative priming because either they think it’s a “wrong result” or because they think it has no value. This is impeding deeper understanding of (rhizosphere) priming effects. Moreover, as it is relatively accepted that priming doesn’t affect the C-balance, the research focus could shift to how priming affects plant nutrition (Possible questions for future research: Has priming the potential to increase plant N-uptake in absence of synthetic fertiliser? Can negative priming and increase of microbial biomass aid soil C sequestration?).
The third chapter claims that rhizosphere priming and general priming are different. Yes, it is important that we include more complexity and understand the rhizosphere effects better. It is a major challenge to study such effects in intact systems that include the whole rhizosphere. This is already discussed in other studies and should be clear. The authors just state that we need to understand rhizosphere priming in natural ecosystems, but they do not provide any way forward to do so. Therefore, I do not see any substantial addition to the discussion based on the very broad and short chapter here.
>> We thank the reviewer for acknowledging the importance of including plants in rhizosphere priming studies. Indeed, addressing the data gap of plant studies in priming research is critical. For example, from the 12 meta-analyses compiled in Xu et al. (2024), n=10 report data exclusively from isolated soil incubations, two included meta-analysis comprise a mix of lab and pot/field experiments (Wang et al., 2016; Feng & Zhu, 2021) and one meta-analysis summarized experiments with living plants (Huo et al., 2017). Wang et al. (2016) is a special case reporting exclusively on PE under biochar and «A few studies were excluded that reported negative values of biochar decomposition because of high variation by application of
C natural abundance or because isotopic fractionation was not considered (Cross & Sohi, 2011)». The meta-meta-analysis also includes a study by Li et al. (2023) reporting PE in response to warming. We have indeed very little information of priming in natural ecosystems and what the possible variability and mechanisms could be under realistic field conditions. We’ll provide more specific suggestions how to improve the current methods and hope this article could stimulate an open debate about priming with a plant perspective.In addition, the authors do not provide any clear definition of the priming effect in the first place. For example, in line 15 it is stated: “Priming effects in soil science describe the influence of labile carbon inputs on rates of microbial soil organic matter mineralisation, which can either increase (positive priming) or decrease (negative priming).” Therefore, they miss the important aspect that priming is affecting the native or present soil organic matter after the addition of new organic matter.
>> Thanks, we’ll clarify this point in the revised manuscript.
Citation: https://doi.org/10.5194/egusphere-2025-1067-AC2 -
AC3: 'Reply on AC2', Jenny Michel, 12 May 2025
We need to correct one sentence: From the 12 meta-analyses compiled in Xu et al. (2024), n=9 report data exclusively from isolated soil incubations, two included meta-analysis comprise a mix of lab and pot/field experiments (Wang et al., 2016; Feng & Zhu, 2021) and one meta-analysis summarized experiments with living plants (Huo et al., 2017).
Citation: https://doi.org/10.5194/egusphere-2025-1067-AC3
-
AC3: 'Reply on AC2', Jenny Michel, 12 May 2025
-
AC2: 'Reply on RC2', Jenny Michel, 12 May 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1067', Anonymous Referee #1, 22 Apr 2025
The paper submitted by Michel et al. explores priming effects in soils, which refer to the influence of labile carbon inputs on the mineralization of organic matter by microbes. These effects can be positive (increased mineralization) or negative (reduced mineralization). The article highlights three key aspects to better understand these effects in their ecological context: (i) Evaluation of net carbon balances (ii) Publication bias and (iii) Difference between general priming effects and rhizosphere-specific priming. The article calls for a more nuanced approach to priming research, encouraging the publication of studies on negative or neutral effects, carefully evaluating publication biases, and distinguishing general priming effects from rhizosphere-specific effects. It also emphasizes the importance of conducting field studies to better understand these phenomena under natural conditions.
I think the points developed by the authors are fair and worth to be published but the current version of the paper is a bit disappointing since it skims over the issues without going into depth. It may come from the format which is not very clear. Is it a review or is it an opinion paper? If this is a review, much more literature must be cited if this an opinion paper, the ideas developed must be more attractive and be inspiring for the community to change their ways of doing priming research.
For instance, the authors wrote that “there is little empirical evidence for net C losses”, I fully agree with the statement but it should be better explained why priming is sometimes see as a mechanism leading to net losses and show some papers that suggest it and explain why they are mistaken. So far the authors mostly cite papers that do not observe any net loss.
The second section is simply a re-analysis of the data from Xu et al, which merely confirms the conclusions of the original paper. In my opinion, this does not add much to the message of Xu et al. More details on the effects of bias should be provided and some ideas to avoid them should be also proposed.
Then the third section is mainly focused on the importance of using intact soils instead of sieved soils whereas the title of the section suggests that RPE and PE should be treated differently. So far this section is not really convincing mostly because it lacks more clear examples and it needs some suggestions to improve the current methods.
Citation: https://doi.org/10.5194/egusphere-2025-1067-RC1 -
AC1: 'Reply on RC1', Jenny Michel, 12 May 2025
Reply to RC1: 'Comment on egusphere-2025-1067', Anonymous Referee #1
The paper submitted by Michel et al. explores priming effects in soils, which refer to the influence of labile carbon inputs on the mineralization of organic matter by microbes. These effects can be positive (increased mineralization) or negative (reduced mineralization). The article highlights three key aspects to better understand these effects in their ecological context: (i) Evaluation of net carbon balances (ii) Publication bias and (iii) Difference between general priming effects and rhizosphere-specific priming. The article calls for a more nuanced approach to priming research, encouraging the publication of studies on negative or neutral effects, carefully evaluating publication biases, and distinguishing general priming effects from rhizosphere-specific effects. It also emphasizes the importance of conducting field studies to better understand these phenomena under natural conditions.
I think the points developed by the authors are fair and worth to be published but the current version of the paper is a bit disappointing since it skims over the issues without going into depth. It may come from the format which is not very clear. Is it a review or is it an opinion paper? If this is a review, much more literature must be cited if this an opinion paper, the ideas developed must be more attractive and be inspiring for the community to change their ways of doing priming research.
>> We thank the reviewer for their time and feedback on this manuscript. We apologies if the format is confusing, it is a “forum article” (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. Purely speculative contributions are discouraged. Manuscript composition is free of parameters, although it is requested that forum articles should be short (less than 2500 words, with a maximum of three figures and tab items-> https://www.soil-journal.net/about/manuscript_types.html). The grains of thought we wish to sow with this article are (i) the increasing evidence that not even positive priming causes net C loss, (ii) the possibility that direction and magnitude of priming effects could be misrepresented due to publication bias and (iii) to scale priming to ecosystem processes methodological changes are needed. The first two points need to be discussed more in the community because the (possibly false) perception of positive priming being a “rule” causes (esp. early career) researchers to repeat and/or discard experiments where they observe negative priming because they think the experiment “didn’t work” or they “made a mistake”. This is a dire situation, and we hope anyone reading the article and having some negative priming data in the corner of their desktop take the time to write it up for peer-reviewed publication because negative priming is a result just as valuable as positive priming. Moreover, if the C-impact of priming is negligible, this means that we can safely consider priming in relation to plant nutrition, which offers many exciting opportunities for future research especially in the field of agronomy. The inspiration the attentive reader could draw from this forum article is therewith two-fold: i) Carefully evaluate their own chain-of-thought and those in the papers they read: Is a net carbon balance provided (this is imperative to talk about net C loss or gain)? Are the hypothesis allowing the option of negative priming occurring? Does the spatial and/or temporal extent over which data is collected match the conclusions / generalizations drawn based on data collected from the study? ii) Carefully think about how future experiments could start with unbiased hypothesis and ideally could include living plants (we'll make some specific suggestions for this in the revised manuscript).
For instance, the authors wrote that “there is little empirical evidence for net C losses”, I fully agree with the statement but it should be better explained why priming is sometimes see as a mechanism leading to net losses and show some papers that suggest it and explain why they are mistaken. So far the authors mostly cite papers that do not observe any net loss.
>> Thanks, we’ll clarify this point in the revised manuscript.
The second section is simply a re-analysis of the data from Xu et al, which merely confirms the conclusions of the original paper. In my opinion, this does not add much to the message of Xu et al. More details on the effects of bias should be provided and some ideas to avoid them should be also proposed.
>> Thanks, we’ll clarify in the manuscript why we think that a meta-meta-analysis with a global average estimate of 10.7% PE (estimated effect size (log-transformed response ratio) of 0.1022 (CI95: 0.0740, 0.1305) is too close to zero priming to state a global rule about the direction of priming given that none of the underlying meta-analysis has been corrected for publication bias. Our re-analysis dropped the author’s priming estimate of +37% to +10.7%, which is 2/3 less than the original estimate (an over-estimation of 60%). It is a huge difference! If we could do a similar exercise of bias correction on the 12 original meta-analysis, it seems possible to very likely that the result would be a global negative to no priming estimate. We would have loved to do this analysis, but unfortunately not all data from the primary meta-analysis was available upon reasonable request. (We were particularly curious about the meta-analysis by Li et al. (2023) which reports exclusively negative priming in response to temperature change but appears with all positive effect sizes in Xu et al.. We would have also loved to better understand how the meta-meta-analysis accounted for the re-appearance of certain primary research papers across the different meta-analysis. For example, Luo et al. 2016 contains n=37 primary studies and n=30 of them re-appear in Sun et al., 2019 and n=43 of the Sun studies re-appear in Yan et al., 2023. The supplementary material of Xu’s meta-meta-analysis includes a redundancy matrix which gives the number of overlap of primary studies included between meta-analysis, but it is not entirely clear how this has been accounted for in the effect sizes (percent change) provided.) We’ll happily dive a bit more into biases of the priming debate in the revised manuscript.
Then the third section is mainly focused on the importance of using intact soils instead of sieved soils whereas the title of the section suggests that RPE and PE should be treated differently. So far this section is not really convincing mostly because it lacks more clear examples and it needs some suggestions to improve the current methods.
>> Thanks, we’ll clarify these points in the revised manuscript. The main point is that many rhizosphere priming studies are not even including a rhizosphere. Sieving soils indeed may also be a problem, not only because it fundamentally changes water hydraulics and therewith plant nutrient uptake (and therewith priming), but also because it possibly affects fungi more than bacteria as the usually used 2mm sieve destroys fungal hyphae but keeps bacteria cells intact. We have some suggestions to improve the current methods, but the best ideas could probably come from an open debate, which is the primary goal of this forum article.
Citation: https://doi.org/10.5194/egusphere-2025-1067-AC1
-
AC1: 'Reply on RC1', Jenny Michel, 12 May 2025
-
RC2: 'Comment on egusphere-2025-1067', Anonymous Referee #2, 24 Apr 2025
In their manuscript with a promising title “What if publication bias is the rule and net carbon loss from priming the exception?” the authors aim to address an important aspect of the relevance of priming. The main point is that priming (i.e. enhanced or reduced mineralization of native soil organic carbon) occurs as an elemental process but the net carbon balance is in most cases not considers in the literature. I agree that this causes a biased understanding of the priming in relation to total carbon changes of soils. In fact, I agree that priming, especially positive, is not on relevant magnitudes to off-set the actual input that is required to initiate priming in the first place. I also generally agree that our understanding is biased due to studies that used “artificial” organic matter inputs and that we have a limited understanding of the complex effects of the whole rhizosphere effect and associated organic matter inputs but also biogeochemical shifts. Overall, I agree that such statements need to be made clearly in order to facilitate future research focuses. However, I have to be honest that I was a little disappointed that the authors remain on a very broad level with their discussion. In addition, it is not clear if this is supposed to be a review or opinion and I miss any clear statement how we should move on.
The first chapter summarizes very briefly that positive priming is not affecting the net carbon balance.
The second chapter re-analysis the study presented by Qin et al (2024) and Xu et al. (2024). In fact, the authors repeat the statement by Qin et al (2024) who stated that the net balance is mainly positive. Therefor, this is supporting the authors point but not providing new information that is shown in Figure 1a that only shows the included studies in Qin et al. Regarding the Xu et al. (2024) study, the chapter in the manuscript is quite technical how the authors re assessed the data and what funnel plots are (line 93-123, which is a lot regarding the short manuscript). The authors provide an improved interpretation of the same plot that is discussed by Xu et al in the supplement figure 2. The authors apply the standard error on ln transformed the priming effect observed in the studies. Technically this seems correct. However, this seems to be rather a response to the Xu et al paper than making a new point. In fact, the Xu et al in the supplement figure 2 already shows the asymmetry that most studies are reporting positive priming effects, when I visually assess the deviation from x=0 in the plot (see plot below). The authors finally estimate that only six studies with moderate or negative priming effects are need to overcome the publication bias. They state in line 137 “It would be interesting to recalculate a global PE estimate from the primary research data of all underlying meta-analysis corrected for publication bias.”. This raises the question why the authors are not providing this. Finally, the authors just encourage to publish negative and moderate priming effects.
The third chapter claims that rhizosphere priming and general priming are different. Yes, it is important that we include more complexity and understand the rhizosphere effects better. It is a major challenge to study such effects in intact systems that include the whole rhizosphere. This is already discussed in other studies and should be clear. The authors just state that we need to understand rhizosphere priming in natural ecosystems, but they do not provide any way forward to do so. Therefore, I do not see any substantial addition to the discussion based on the very broad and short chapter here.
In addition, the authors do not provide any clear definition of the priming effect in the first place. For example, in line 15 it is stated: “Priming effects in soil science describe the influence of labile carbon inputs on rates of microbial soil organic matter mineralisation, which can either increase (positive priming) or decrease (negative priming).” Therefore, they miss the important aspect that priming is affecting the native or present soil organic matter after the addition of new organic matter.
Funnel plot from Xu et al (2024, supplement) with added x=0 vertical line
Citation: https://doi.org/10.5194/egusphere-2025-1067-RC2 -
AC2: 'Reply on RC2', Jenny Michel, 12 May 2025
Reply to RC2: 'Comment on egusphere-2025-1067', Anonymous Referee #2
In their manuscript with a promising title “What if publication bias is the rule and net carbon loss from priming the exception?” the authors aim to address an important aspect of the relevance of priming. The main point is that priming (i.e. enhanced or reduced mineralization of native soil organic carbon) occurs as an elemental process but the net carbon balance is in most cases not considers in the literature. I agree that this causes a biased understanding of the priming in relation to total carbon changes of soils. In fact, I agree that priming, especially positive, is not on relevant magnitudes to off-set the actual input that is required to initiate priming in the first place. I also generally agree that our understanding is biased due to studies that used “artificial” organic matter inputs and that we have a limited understanding of the complex effects of the whole rhizosphere effect and associated organic matter inputs but also biogeochemical shifts. Overall, I agree that such statements need to be made clearly in order to facilitate future research focuses.
>> We thank the reviewer for their time and feedback on this manuscript. We apologise for the brevity and confusion about the “forum” format and will be more specific about possible avenues of future research (which requires that we critically evaluate the conclusions of past and present research which is neglecting negative priming).
However, I have to be honest that I was a little disappointed that the authors remain on a very broad level with their discussion. In addition, it is not clear if this is supposed to be a review or opinion and I miss any clear statement how we should move on.
>> Thanks, reviewer 1 said the same thing.
The first chapter summarizes very briefly that positive priming is not affecting the net carbon balance.
>> Yes.
The second chapter re-analysis the study presented by Qin et al (2024) and Xu et al. (2024). In fact, the authors repeat the statement by Qin et al (2024) who stated that the net balance is mainly positive. Therefor, this is supporting the authors point but not providing new information that is shown in Figure 1a that only shows the included studies in Qin et al. Regarding the Xu et al. (2024) study, the chapter in the manuscript is quite technical how the authors re assessed the data and what funnel plots are (line 93-123, which is a lot regarding the short manuscript). The authors provide an improved interpretation of the same plot that is discussed by Xu et al in the supplement figure 2. The authors apply the standard error on ln transformed the priming effect observed in the studies. Technically this seems correct. However, this seems to be rather a response to the Xu et al paper than making a new point. In fact, the Xu et al in the supplement figure 2 already shows the asymmetry that most studies are reporting positive priming effects, when I visually assess the deviation from x=0 in the plot (see plot below). The authors finally estimate that only six studies with moderate or negative priming effects are need to overcome the publication bias. They state in line 137 “It would be interesting to recalculate a global PE estimate from the primary research data of all underlying meta-analysis corrected for publication bias.”. This raises the question why the authors are not providing this. Finally, the authors just encourage to publish negative and moderate priming effects.
>> We thank the reviewer for acknowledging the improvements our re-analysis brings to the interpretation. We would have been delighted to re-analyse the underlying meta-analysis, including the C-in vs. out-puts, study duplication, publication bias etc. but the data was not available for all articles. We are not sure if it is realised in the community how strong the impact of confirmation and conformity bias is? The publication of negative and no priming effects needs to be encouraged strongly because some researchers do not publish negative priming because either they think it’s a “wrong result” or because they think it has no value. This is impeding deeper understanding of (rhizosphere) priming effects. Moreover, as it is relatively accepted that priming doesn’t affect the C-balance, the research focus could shift to how priming affects plant nutrition (Possible questions for future research: Has priming the potential to increase plant N-uptake in absence of synthetic fertiliser? Can negative priming and increase of microbial biomass aid soil C sequestration?).
The third chapter claims that rhizosphere priming and general priming are different. Yes, it is important that we include more complexity and understand the rhizosphere effects better. It is a major challenge to study such effects in intact systems that include the whole rhizosphere. This is already discussed in other studies and should be clear. The authors just state that we need to understand rhizosphere priming in natural ecosystems, but they do not provide any way forward to do so. Therefore, I do not see any substantial addition to the discussion based on the very broad and short chapter here.
>> We thank the reviewer for acknowledging the importance of including plants in rhizosphere priming studies. Indeed, addressing the data gap of plant studies in priming research is critical. For example, from the 12 meta-analyses compiled in Xu et al. (2024), n=10 report data exclusively from isolated soil incubations, two included meta-analysis comprise a mix of lab and pot/field experiments (Wang et al., 2016; Feng & Zhu, 2021) and one meta-analysis summarized experiments with living plants (Huo et al., 2017). Wang et al. (2016) is a special case reporting exclusively on PE under biochar and «A few studies were excluded that reported negative values of biochar decomposition because of high variation by application of
C natural abundance or because isotopic fractionation was not considered (Cross & Sohi, 2011)». The meta-meta-analysis also includes a study by Li et al. (2023) reporting PE in response to warming. We have indeed very little information of priming in natural ecosystems and what the possible variability and mechanisms could be under realistic field conditions. We’ll provide more specific suggestions how to improve the current methods and hope this article could stimulate an open debate about priming with a plant perspective.In addition, the authors do not provide any clear definition of the priming effect in the first place. For example, in line 15 it is stated: “Priming effects in soil science describe the influence of labile carbon inputs on rates of microbial soil organic matter mineralisation, which can either increase (positive priming) or decrease (negative priming).” Therefore, they miss the important aspect that priming is affecting the native or present soil organic matter after the addition of new organic matter.
>> Thanks, we’ll clarify this point in the revised manuscript.
Citation: https://doi.org/10.5194/egusphere-2025-1067-AC2 -
AC3: 'Reply on AC2', Jenny Michel, 12 May 2025
We need to correct one sentence: From the 12 meta-analyses compiled in Xu et al. (2024), n=9 report data exclusively from isolated soil incubations, two included meta-analysis comprise a mix of lab and pot/field experiments (Wang et al., 2016; Feng & Zhu, 2021) and one meta-analysis summarized experiments with living plants (Huo et al., 2017).
Citation: https://doi.org/10.5194/egusphere-2025-1067-AC3
-
AC3: 'Reply on AC2', Jenny Michel, 12 May 2025
-
AC2: 'Reply on RC2', Jenny Michel, 12 May 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 984 | 66 | 22 | 1,072 | 34 | 57 |
- HTML: 984
- PDF: 66
- XML: 22
- Total: 1,072
- BibTeX: 34
- EndNote: 57
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Jennifer Michel
Yves Brostaux
Bernard Longdoz
Hervé Vanderschuren
Pierre Delaplace
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(612 KB) - Metadata XML
The paper submitted by Michel et al. explores priming effects in soils, which refer to the influence of labile carbon inputs on the mineralization of organic matter by microbes. These effects can be positive (increased mineralization) or negative (reduced mineralization). The article highlights three key aspects to better understand these effects in their ecological context: (i) Evaluation of net carbon balances (ii) Publication bias and (iii) Difference between general priming effects and rhizosphere-specific priming. The article calls for a more nuanced approach to priming research, encouraging the publication of studies on negative or neutral effects, carefully evaluating publication biases, and distinguishing general priming effects from rhizosphere-specific effects. It also emphasizes the importance of conducting field studies to better understand these phenomena under natural conditions.
I think the points developed by the authors are fair and worth to be published but the current version of the paper is a bit disappointing since it skims over the issues without going into depth. It may come from the format which is not very clear. Is it a review or is it an opinion paper? If this is a review, much more literature must be cited if this an opinion paper, the ideas developed must be more attractive and be inspiring for the community to change their ways of doing priming research.
For instance, the authors wrote that “there is little empirical evidence for net C losses”, I fully agree with the statement but it should be better explained why priming is sometimes see as a mechanism leading to net losses and show some papers that suggest it and explain why they are mistaken. So far the authors mostly cite papers that do not observe any net loss.
The second section is simply a re-analysis of the data from Xu et al, which merely confirms the conclusions of the original paper. In my opinion, this does not add much to the message of Xu et al. More details on the effects of bias should be provided and some ideas to avoid them should be also proposed.
Then the third section is mainly focused on the importance of using intact soils instead of sieved soils whereas the title of the section suggests that RPE and PE should be treated differently. So far this section is not really convincing mostly because it lacks more clear examples and it needs some suggestions to improve the current methods.