Use of soil respiration measurements and RothC modelling show effects of catch crops and precision and traditional agriculture on productivity and soil organic carbon dynamics in a 5 year study in Mediterranean climate

Balugani, Enrico; Castellucci, Alessia; Ruggeri, Matteo; Meriggi, Pierluigi; Volta, Benedetta; Marazza, Diego

doi:https://doi.org/10.5194/egusphere-2023-2966

Preprints

https://doi.org/10.5194/egusphere-2023-2966

Preprints

22 Dec 2023

| 22 Dec 2023

Use of soil respiration measurements and RothC modelling show effects of catch crops and precision and traditional agriculture on productivity and soil organic carbon dynamics in a 5 year study in Mediterranean climate

Enrico Balugani, Alessia Castellucci, Matteo Ruggeri, Pierluigi Meriggi, Benedetta Volta, and Diego Marazza

Abstract. Finding agricultural managements able to increase soil organic carbon without a reduction in crop yields is important to: decrease soil erosion, protect soil ecosystem services, increase soil health, help to curb net CO₂ emissions toward the EU goal of carbon neutrality. Various studies have shown that catch crops, when managed in the proper way, may result in an increase in soil carbon stocks; however, recent studies have cast doubts on those findings, due to short study duration (3 years or less), few data points, and catch crops mismanagement. Model studies to estimate the potentials of catch crops for soil carbon sequestration shown mixed results; however, in these studies, only the direct effects of catch crops (i.e. the input of carbon from crop inclusion in the soil) was accounted for. Here, we show the result of a study to compare two crop managements: traditional against catch crop together with precision agriculture. We measured agricultural productivity, soil organic carbon, soil respiration, and soil conditions in two different sites in Italy for a period of 4+ years, then we modelled the field managements using a modified version of RothC model, to account for both direct and indirect catch crop effects on soil. The results show that catch crops and precision agriculture can result in an increase in soil organic carbon, with no effects, or, in some cases, an increase in crop production.

Received: 08 Dec 2023 – Discussion started: 22 Dec 2023

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Enrico Balugani, Alessia Castellucci, Matteo Ruggeri, Pierluigi Meriggi, Benedetta Volta, and Diego Marazza

Status: final response (author comments only)

RC1:
'Comment on egusphere-2023-2966', Anonymous Referee #1, 19 Jan 2024
General comment
This manuscript aims at answering several very interesting questions, including: (1) the effect of catch crops on soil organic carbon stocks and agricultural productivity, (2) comparing an “efficient cropping system” to a “conventional cropping system” where not only the crop rotations but also other management measures (such as fertilization and irrigation) are different, (3) multi-objective calibration of the RothC model to catch crops in the Mediterranean region, and (4) testing and comparing the ability of two RothC versions to simulate SOC dynamics of catch crops in the Mediterranean (i.e., default RothC vs. a modified version of RothC called “RothC20_N” that is calibrated for arid conditions in the Mediterranean region). They suggest interesting and valid approaches to some of these questions.
Unfortunately, the authors switch the topic several times during the different sections of the manuscript without keeping a clear structure. The main message and scope of the manuscript remains unclear. Large parts of the material&methods and the results sections are seemingly missing since the discussion and conclusion refer to results or methods that are not shown in the corresponding sections. This makes it difficult to judge on the scientific quality. The approaches and results are not presented in a way that would answer the several (indirect, there are no direct) research questions.
However, since both the data-sets and modelling approaches seem promising and interesting, I would like to encourage the authors to concentrate on a main topic and keep a revised manuscript as close to this main topic as possible. From what I read, I suggest that this could be (1) comparing two cropping systems, and (2) testing and calibrating the RothC model to simulate these cropping systems. Here are my specific comments regarding the current manuscript:
Major specific comments
The treatments “efficient cropping system” (ECS) and “conventional cropping system” (CCS) are not described clear enough. Differences between the treatments are not shown or described in a quantitative way. In large parts of the manuscript, the authors write as if they see the ECS as the “catch crop treatment” and CCS as the reference and as if growing catch crops was the only difference between the cropping systems. However, they also indicate that there are differences in the fertilization rates, irrigation rates and crop rotations and other cropland managements, but without quantifying and showing these. This would mean that this field experiment is not a comparison of “cover crops” vs “bare fallow” where everything else is comparable, but rather a comparison of different cropping systems. Observed yield increases and SOC increases would thus not be a result of catch crops alone, but a result of combined measures: optimized irrigation, catch crops, different crop rotation in general etc.

The research questions are not clear. From the abstract, the reader could think that the topic is carbon sequestration effect of catch crops. However, the last paragraph of the intro indicates that the main objective is calibrating RothC or testing RothC’s ability to describe catch crop effects in Mediterranean climate. In the M&M part, it becomes clear that it’s more a comparison of cropping systems, and a new method to separate respiration rates in order to calibrate RothC. However, more specific information on all these questions is missing. For example: How exactly was precision agriculture implemented here (it is only briefly mentioned in the title, introduction and conclusion)? What about the irrigation and fertilization (rates and types)? How were roots and root respiration estimated?

The abstract structure needs revision: The introductory part of the abstract is long; the results are only one sentence and there is no concluding sentence at all. Again, the main research question and message is missing.

The authors mention and discuss a questionable study of Chaplot & Smith (2023) several times. A recent letter to that paper (https://doi.org/10.1111/gcb.17128) suggests that the "cherry-picking" approach of that paper is unscientific, that the studies were misinterpreted and that the results are quite weak. Instead, there is overwhelming scientific evidence of an average positive effect of cover crops on SOC stocks based on hundreds of field studies. Thus, it might not be necessary to elaborate on that paper too much. Apart from this, I see this question out of the original scope of this manuscript and therefore suggest to concentrate on those aspects the field and modelling study was originally set up for. From my understanding, the experimental setup does not allow to verify the solely cover crop effect on SOC and yield (see point 1).

Minor specific comments
Some specific comments on where I see missing information, missing results and room for improved presentation:
What were the parameter estimates (l. 328)?

How did the authors quantify or decide which model fit was best (l. 308, l. 325)?

What was the rooting depth (l. 312) and, in general, how were the roots observed/measured/estimated (l. 246) and what were the results (l. 430)?

What was the estimated carbon input (l. 352)?

In l. 379, only “less CO2 emissions” are quantified, the other factors are not quantified or shown, but sometimes mentioned (e.g., l. 445). Please include these results in the results section.

What does 4+ years mean? The title says 5 years, the body says 4+ years (e.g. in l. 96), and there are only results for 3 years (Fig. 3, Fig. 4).

Please decide for one name: “catch crops” or “cover crops” or even “relay crops” (l. 441)? Keep it consistent.

Please check the resolution of the figures.

Please consider to structure the discussion section using subsections.
Citation: https://doi.org/10.5194/egusphere-2023-2966-RC1
- AC2: 'Reply on RC1', Enrico Balugani, 23 Mar 2024
  
  We thank the reviewer for the kind words about our Manuscript. We agree that there may be too many different topics in the study. The idea was to include these topics in the manuscript to show all the different aspects of the Life EU project. However, We agree that this leads to an unfocused manuscript. Therefore, after reading the opinions of both Reviewers, We decided to focus more on the testing and calibrating of the RothC model for the specific conditions (two different practices, in a Mediterranean area).
  
  Due to this refocusing of the objectives, We will restructure the manuscript to address the research question "can RothC identify the carbon dynamics due to two different practices [...]", keeping to a minimum the more "agronomical" parts of the manuscript dealing with the differences between the CCS and ECS managements, and including more information on the modeling itself. The abstract will be revised accordingly.
  Answering to the major points one by one:
  
  1. We will include more information on the two practices (ECS, CCS), however, since We decided to focus more on the modeling aspects, We will probably provide them in the supplementary material, to avoid confusion.
  
  2. As described in the general answer above, we will restructure the manuscript focusing on the modeling part, and the abstract accordingly.
  
  3. We will also structure the abstract with less introduction and more methods, results and discussion.
  
  4. We decided to include a couple of paragraphs to show that this study "passes the requirements" defined in Chaplot & Smith (2023), due to a previous, unrelated manuscript review experience. However, We agree that the paragraphs can be taken out of the revised version of the manuscript.
  Minor specific comments:
  
  1. We will give the whole list of calibrated parameters, however We calibrated only one parameter for the water balance (using the dataset of soil water content obtained from the field) and the degradation constants Khum and krpm, which were the most important ones for the time scale considered.
  2. We will give a metric for the decision of best model fit; we used the R-squared and Nash-Sutcliffe efficiency coefficient.
  3. The rooting depth depends on the crop grown, it goes from 30 cm (Pea) to 1 m (Maize). The root growth comes from the crop model used by the Decision Support System (the same used to determine crop management in the ECS), which is based on direct observations taken through the years by Hort@ s.r.l. We will include the root growth estimates in the revised version of the supplementary materials.
  4. The carbon input was estimated in two ways: the above ground biomass left on the field after harvest was directly measured, while the below ground biomass comes from the physiological models of the DSS, calibrated on direct observations of root biomass (see answer above) .
  5. the results will be included in the reviewed manuscript.
  6. The experiment started on 1st January 2018, and is still going on. However, the Life project started only in 2020. Therefore, for the first year (2018) we have only the total amount of above ground biomass production; for 2019 we have more data, but the gas chambers and the soil sensors were only installed when the funding was available, in 2020.
  7. We will use consistently the "catch crop" label.
  8. Most figure are prepared in Python, and We decided to keep 150 dpi for the first manuscript, and to shift to 300 dpi for the final version. We will provide the 300 dpi version of the figure in the revised manuscript.
  9. We will divide the discussion section into subsections, probably including "discussion of the results"; "comparison with other studies", "limitations of the study".
  
  Citation: https://doi.org/10.5194/egusphere-2023-2966-AC2
AC1: 'Comment on egusphere-2023-2966', Enrico Balugani, 13 Mar 2024

I attach here the Supplementary materials, previously not provided due to decisions to be made upon disclosure of data owned by Hort@ s.r.l.

Citation: https://doi.org/10.5194/egusphere-2023-2966-AC1
RC2:
'Comment on egusphere-2023-2966', Anonymous Referee #2, 13 Mar 2024
“Use of soil respiration measurements and RothC modelling show effects of catch crops and precision and traditional agriculture on productivity and soil organic carbon dynamics in a 5 year study in Mediterranean climate”
This manuscript is moderately well written. It applies a useful, potentially reproducible approach to calibrate process-based soil organic carbon models based on a combination of measurements (fast rate of measurement, low uncertainty (CO2) and slow rate of measurement, large uncertainty (SOC). It furthermore evaluates the effect of cover crops and precision agriculture on SOC dynamics and can represent a valuable contribution to evaluating climate-friendly agricultural practices.
On the other hand, the methodology and results section require clarification, particularly in relation to the modelling approach and statistical evaluation. Additionally, the systems (CSS/ECS) are not well defined, particularly in regard to precision agriculture. Furthermore, attributing the yield- and SOC-enhancing effects to cover crops or precision agriculture is difficult, as it is a combined measure.
My suggestions are detailed as follows:
Title
The title is too convoluted and does not reflect the main topic of the manuscript well. The title suggests a relation between soil respiration/RothC modeling and agricultural productivity, yet no such relation is described in the manuscript nor does RothC contain a plant-growth sub-model which could help support that claim. Therefore, the title could be adapted to “Use of soil respiration measurements and RothC modelling show effects of catch crops and precision and traditional agriculture on soil organic carbon dynamics in a 5-year study in Mediterranean climate”.
Introduction
L68: “Over-parametrization problem can lead to equifinality” was overfitting meant instead of equifinality? If not, a short description of equifinality would help the reader.
L110: Please recheck that the Foggia site in classified as BSk by Köppen-Geiger.
Materials and methods
2.2: Suggest to describe the management resulting from the DSS, as it is not clear what the management includes.
L120: Suggest to use ”conservation tillage” instead of conservative soil tillage.
L129,L130: Fig. 2 does not show higher N availability or an increase in total productivity, it is assumed, but not tested.
2.3: I would only mention measurements that were used in the study.
L141: How was it tested that field 4 & 5 are representative for ECS/CCS conditions.
L150: Suggest to change “humidity” to water content.
L157: Why was Walkley & Black used for SOC determination?
L189: Parameter fits are described to be obtained within R in the manuscript and Python in the supplementary materials. Please revise.
L196: mentions that there is no sub-model for the simulation of plant growth, contradicting the title, which implies that RothC modelling shows the effects of cover crops and precision agriculture on productivity and SOC.Please adapt.
L244: Here it mentions that bulk density was measured in the field, while in L 157 it mentions an estimated BD. From the supplementary material we can see that it was estimated, as it is the same for all sites and treatments.Please adapt.
In the methods section, there are two important components missing:
Statistical evaluation: there is some mentions of significant/non-significant trends etc, but no description of statistical tests.

Model evaluation criteria: When comparing models and evaluating their predictive performance, model performance metrics should be used to justify selecting one model over another.

Results
As the partition of soil respiration is a central topic of this manuscript, I would advise to include the results in the text and a graph showing the relation between respiration and root biomass. Statistical tests should have been defined in the previous section, and differences between managements should be tested. Some plots (e.g. S4 and S7) from the supplementary materials can be included in the manuscript, but require statistical testing metrics, which are yet to be provided. Please revise.
L281: “No clear difference visible” why was it not tested statistically?
L298: Which fertilizer does this refer to?
L306: fitting less satisfactory based on what metric? Please elaborate.
L310: Residuals refer to the same quantity, i.e. observed and predicted should be the same (either Rh or Rs)
L355: Which test was used?
L338: Same as before.
L360: Plots 4 and 5 were used in the calibration of the model, it should not be used for model evaluation.
L364: Kdpm is 20% larger than the original, not < 5%.
L381: data about profits is not shown or discussed, suggest removing the text or included the data.
L428: Here it says 8+ years are required to calibrate a model, while it says +10 in other sections. Please adapt.
Overall the manuscript has good potential but I suggest that it requires a major revision to be eligible for publication.
Citation: https://doi.org/10.5194/egusphere-2023-2966-RC2
- AC3: 'Reply on RC2', Enrico Balugani, 23 Mar 2024
  
  We thank the reviewer for the comments and suggestions, and for the kind opinion on our manuscript.
  The problems identified by Reviewer 2 go in the same direction of Reviewer 1: We included too many topics in the manuscript, making it confusing and untidy. We decided to focus on the modelling approach only, as per the answer to Reviewer 1, and therefore cut most of the material which is not focused on the modelling, while including more detailed information on the modelling itself, including a metric for model performance (R-squared and Nash-Sutcliffe efficiency coefficient).
  To provide a point-by-point response:
  Title:
  - We accept the proposition of the Reviewer, We will only adapt the title to the more focused objective of the manuscript (the modelling).
  
  Introduction:
  - We will keep only overparametrization in the introduction. However, there is also the problem of equifinality (different sets of parameters and inputs leading to the same results) - but We prefer to make the introduction lighter and more focused.
  
  Materials and methods:
  - we will provide more information on the managements, however, since the focus will shift more onto the modelling, We will provide the information in the supplementary materials.
  
  - plot 4 and 5 have the same soil material as the whole area, the same management in the first 2 years, the same biomass production as the other plots with the same management (either CCS or ECS). This was the same reason for considering plot 4 and 5 representative for the related managements.
  
  - We used Walkley & Black because We had no other method available at the time (the analyses were conducted by the agricultural company Hort@ s.r.l.).
  
  - The bulk density has been measured every year. However, We noted that the measurements were taken after the soil was "worked", as such, we decided to use, instead of the direct measurements, estimates coming from the measurements taken at the beginning of the experiment, when the soil was in "standard" conditions.
  
  Results:
  - we will include metrics for model fit performance, and we will include the statistical analyses used.
  
  - We will also include more information on the soil respiration partitioning, which now is in supplementary materials, in the main manuscript.
  
  - L298: We should have written "fertilizers", since it refers to the whole amount of fertilizers used in the 5 years of management. We will include a table with information on fertilizers used and related amount, for each year of experiment.
  
  - L381: due to the "refocusing" of the manuscript, all this information about comparison between the two managements will be taken out of the main manuscript.
  
  all the other minor comments will be accepted in the revised manuscript.
  
  Citation: https://doi.org/10.5194/egusphere-2023-2966-AC3

Enrico Balugani, Alessia Castellucci, Matteo Ruggeri, Pierluigi Meriggi, Benedetta Volta, and Diego Marazza

Viewed

Total article views: 649 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
450	162	37	649	25	29

HTML: 450
PDF: 162
XML: 37
Total: 649
BibTeX: 25
EndNote: 29

Views and downloads (calculated since 22 Dec 2023)

Month	HTML	PDF	XML	Total
Dec 2023	84	23	9	116
Jan 2024	57	13	3	73
Feb 2024	32	10	1	43
Mar 2024	71	32	7	110
Apr 2024	38	17	4	59
May 2024	28	20	3	51
Jun 2024	31	13	2	46
Jul 2024	50	14	5	69
Aug 2024	28	2	2	32
Sep 2024	10	4	1	15
Oct 2024	11	8	0	19
Nov 2024	10	6	0	16

Cumulative views and downloads (calculated since 22 Dec 2023)

Month	HTML	PDF	XML	Total
Dec 2023	84	23	9	116
Jan 2024	57	13	3	73
Feb 2024	32	10	1	43
Mar 2024	71	32	7	110
Apr 2024	38	17	4	59
May 2024	28	20	3	51
Jun 2024	31	13	2	46
Jul 2024	50	14	5	69
Aug 2024	28	2	2	32
Sep 2024	10	4	1	15
Oct 2024	11	8	0	19
Nov 2024	10	6	0	16

Viewed (geographical distribution)

Total article views: 669 (including HTML, PDF, and XML) Thereof 669 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 20 Nov 2024

Short summary

To become more sustainable, agriculture needs practices that decrease net carbon emission to the atmosphere, increase soil carbon, and simultaneously increase yields and profits. In this study, we show that, in two different areas in Italy, the use of technology-informed managements of catch and cover crops could increase soil carbon sequestration, decrease overall carbon emissions, and increase agricultural yields.


Total:	0
HTML:	0
PDF:	0
XML:	0