Effect of terrestrial nutrient limitation on the estimation of the remaining carbon budget

De Sisto, Makcim L.; MacDougall, Andrew H.

doi:https://doi.org/10.5194/bg-2023-96

Preprints

https://doi.org/10.5194/bg-2023-96

Preprints

04 Jul 2023

| 04 Jul 2023

Status: a revised version of this preprint is currently under review for the journal BG.

Effect of terrestrial nutrient limitation on the estimation of the remaining carbon budget

Makcim L. De Sisto and Andrew H. MacDougall

Abstract. The carbon cycle plays a foundational role in the estimation of the remaining carbon budget. It is intrinsic for the determination of the transient climate response to cumulative CO₂ emissions and the zero emissions commitment. For the terrestrial carbon cycle, nutrient limitation has a core regulation on the amount of carbon fixed by terrestrial vegetation. Hence, the addition of nutrients such as nitrogen and phosphorus in land model structures in Earth system models is essential for an accurate representation of the carbon cycle feedback in future climate projections. Thereby, the estimation of the remaining carbon budget is impacted by the representation of nutrient limitation in modelled terrestrial ecosystems, yet it is rarely accounted for. Here, we estimate the carbon budget and remaining carbon budget of a nutrient limited Earth system model, using nitrogen and phosphorus cycles to limit vegetation productivity and biomass. We use eight Shared Socioeconomic Pathways scenarios and idealized experiments on three distinct model structures: 1) carbon cycle without nutrient limitation, 2) carbon cycle with terrestrial nitrogen limitation and 3) carbon cycle with terrestrial nitrogen and phosphorus limitation. To capture the uncertainty of the remaining carbon budget, three different climate sensitives were tuned for each model version. Our results show that overall the nutrient limitation reduced the remaining carbon budget for all simulations in comparison with the carbon cycle without nutrient limitation. Between the nitrogen and nitrogen-phosphorus limitation, the latter had the lowest remaining carbon budget. The mean remaining carbon budget from the Shared Socioeconomic Pathways scenarios simulations for the 1.5 °C target in the no nutrient limitation, nitrogen limited and nitrogen-phosphorus limited models obtained were 228, 185 and 175 Pg C respectively, relative to year 2020. For the 2 °C target the mean remaining carbon budget were 471, 373 and 351 Pg C for the no nutrient limitation, nitrogen limited and nitrogen-phosphorus limited models respectively, relative to year 2020. This represents a reduction of 19 and 24 % for the 1.5 °C target and 21 and 26 % for the 2 °C target in the nitrogen and nitrogen-phosphorus limited simulations compared to the no nutrient limitation model. These results show that terrestrial nutrient limitations constitute an important factor to be considered when estimating or interpreting remaining carbon budgets and are an essential uncertainty of remaining carbon budgets from Earth system model simulations.

Received: 02 Jun 2023 – Discussion started: 04 Jul 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

Makcim L. De Sisto and Andrew H. MacDougall

Status: final response (author comments only)

RC1:
'Comment on bg-2023-96', Anonymous Referee #1, 21 Jul 2023

The manuscript shows the impact of nutrient limitation on estimating the remaining carbon budget in Earth system models. The C cycle and its interaction with nutrients play a crucial role in determining the amount of C that terrestrial vegetation can take up. However, the representation of nutrient cycles in modeled terrestrial ecosystems is often overlooked, despite its influence on the C cycle feedback in future climate projections. The study estimates the C budget and the remaining C budget using an Earth system model (UVic ESCM) that incorporates N and P cycles to limit vegetation productivity and biomass. The results demonstrate that nutrient limitation reduces the remaining C budget in comparison to models without nutrient limitation, with the N-P limitation resulting in the lowest remaining carbon budget. The findings highlight the significance of considering nutrient limitations when estimating and interpreting remaining C budgets, highlighting the importance of accurately representing these limitations in Earth system model simulations.

This research topic is very important due to its implications for understanding the C cycle and predicting future climate change. Accurately estimating the remaining C budget is key for developing effective climate mitigation strategies. By incorporating nutrient cycles into ESMs, this study shows that neglecting to account for nutrient limitations can lead to uncertain projections of C uptake and release by terrestrial vegetation. Therefore, understanding the role of nutrient limitation in the C cycle and considering it in modeling frameworks is crucial for improving the accuracy of climate change predictions and guiding appropriate mitigation efforts.

The manuscript would benefit from an overall improvement in writing for clarity. References must be checked, as I have found at least one missing reference; the use of a reference manager is advised. Moreover, often, the source of a method or approach is mentioned in the method's section without further explanation. It would be helpful for the reader to provide a short explanation of the method/approach or some context. Specifically, the description of the nutrient cycles, which are the focal point of the manuscript, could be highly improved so the reader does not have to go for a second publication to learn the basics about what the model is doing. Also important, the manuscript lacks a more robust discussion on the uncertainties of the assumptions of the model. What is missing, and what could be improved in the future? So the reader can properly assess and interpret the results.

Based on the above, I would not recommend the publication of this manuscript without further major revisions.

See specific comments below.

Line 9: Add SSPs between parentheses to introduce the acronym.

Line 30: Missing reference at the end of the sentence. "Since year 1850, the cumulative CO2 land sink has been estimated to be 210±45 PgC, which represents 31% of all anthropogenic carbon emissions".

Lines 58: Consider reading Fleischer et al., 2019, which might be a good reference there.

Very long first paragraph in the introduction. Consider splitting it.

Lines 69-70: "The remaining carbon budget framework used in this study follows"… Sounds like a sentence that should be in the methods and could be rephrased.

Line 94: "budgets budgets" appears repeatedly along the manuscript. Please clarify.

Line 121: What does the process of biological nitrogen fixation and mineralization of organic nitrogen depend on in the model?

Lines 133-134: "The soil litter decomposed is transferred to the soil organic P pool. The dynamics of P organic matter are adapted from Wang et al. (2007)." It would be helpful for the reader to provide a short description of the dynamics. The same is true for line 130, where Goll et al. (2017) are cited. What do these dynamics depend on?

How are N and P together influencing SOM decomposition in the model?

Description of the terrestrial model: The methods do not mention leaf nutrient resorption. That's an important aspect of nutrient demand. Although the full description of the model can be found elsewhere, a more complete summary of the terrestrial N and P model would be helpful for the reader.

Line 146: The mention of tuning climate sensitivity using an equilibrium climate sensitivity parameter designed by Zickfeld et al. (2009) could benefit from further explanation. Explaining how the tuning process alters the flow of long-wave radiation back to space and how it impacts the climate sensitivity in the model would improve understanding.

Line 160: The mention of "Tokarska et al. (2018) approach" is not explained. It would be helpful to provide a brief description of the approach.

Line 161: The acronym in the sentence "SSP 5-8.5 a high emission scenario" is introduced without any prior explanation. It would be helpful to provide a brief description of the SSPs scenarios and what the 5-8.5 scenario represents.

Line 165: The term "Zero Emission Commitment Model Intercomparison Project (ZECMIP) protocol" is introduced without prior explanation. It would be beneficial to provide a brief description of what the ZECMIP protocol entails.

Line 166: The statement "the 1pctCO2 experiment is followed until diagnosed cumulative emissions of CO2 reaches 1000 PgC thereafter emissions are set to zero further CO2 emissions" could be clarified. It is not clear what is meant by "diagnosed cumulative emissions". Providing more details on the methodology would improve understanding.

Lines 175-178: The SSP scenarios are again mentioned without explanation, a brief general description of what they mean would be helpful for the readers.

Line 179: Throughout the manuscript, I am confused with how you use the term "land use change." Please, clearly define it. It seems the term is being used to describe natural changes in vegetation cover. But land use change is related to anthropogenic transformations of the natural landscape.

Line 183: punctuation missing

Line 191: Term GISS introduced without further clarification. Additionally, was this mentioned in the methods?

Line 204: I believe the correct wording should be "less carbon is taken up from the land." Please, check the rest of the manuscript for the same issue (e.g., line 220, replace uptake for take up).

Line 213: It would be helpful to remind the reader what the TCR means and how to interpret it. For instance: "the amount of global warming expected to occur when atmospheric CO2 concentrations double from their pre-industrial levels, while all other factors remain constant", or what you think is more appropriate.

Line 218: Sentence could be improved; it is not complete, should be “vegetation uptake of carbon” I believe.

Line 222: Replace: "terrestrial vegetation is constrained by nutrient…"

Line 235: Remind the reader of what is and how to interpret ZEC values. For instance: "values indicate the estimated temperature increase resulting from the ZEC after 50 years of no further emissions. Higher ZEC value suggest… (continue)."

Line 262: Replace "Figures 2-8."

Lines 273-274: "As the model reduces vegetation due to nutrient limitation and trees are replaced by grassed, the land surface albedo is increased." This needs further clarification. In what parts of the world is this happening, and what ecosystems are being replaced by grass? Additionally, replace grassed by grasslands.

Lines 272-274: Here, you are talking about land use change, that trees are being replaced by grasses due to nutrient limitation. Is this land use change? I would say anthropogenic activities, not nutrient limitation, drive land use change. Nutrient limitations might cause natural changes in vegetation cover.

Lines 282-283: "When the diagnosed C:N or C:P leaf ratios are higher than the set maximum leaf ratio, the vegetation biomass dies so that the leaf ratios decrease back to the maximum ratio threshold." This sentence needs further explanation. If there's a maximum set, how can it go higher? I don't fully understand what is the assumption here with the biomass dying and the ratios being reset.

Lines 283-284: First time the acronym PFT is used, it is not defined, and no further clarification is given—the relevant differences among PFTs were not summarized in the methods or there. Also, be consistent with the use of all acronyms throughout the text; either use them or not.

Lines 284: "nutrient limiting stressors such as nitrogen and phosphorus should be applied carefully as a high limitation of phosphorus can easily underestimate the land sink capacity of tropical vegetation". This needs additional clarification. What in the model's representation of the P cycle could be causing P to be too limiting, causing this underestimation? Considering you do not account for the C costs of P uptake, I would expect the other way around. What are the limitations of the P cycle in the model? Couldn't other assumptions in the model, not related to the P cycle, be the problem in tropical regions?

Lines 305-306: "These estimations are larger than the rough estimate of 27 PgC reduction of carbon budgets due to unrepresented carbon feedbacks (Rojeli et al., 2018), suggesting that this value may have been underestimated in the IPCC 1.5◦C report." What did Rojeli et al. do differently? Rojeli is not on the reference list; is that Rogelj? Please check the references and reference list for such mistakes. The use of a reference manager is recommended.

Line 320: Revise the sentence; it is unclear. "Although, different using ECS might assess some variability shown in other models, there is a clear uncertainty in how variable the impacts is in other model structures."

Discussion: the manuscript lacks an important aspect, which is a more robust discussion of the model's limitations. What are the most important limitations and assumptions of the model (Land surface model or otherwise)? What needs to be improved? You briefly speak of the lack of data, some uncertainties, but what uncertainties, and what data specifically? (Starting on line 315).

Line 315: Not clear, nutrient concentration in what? Soil? Leaf? Biomass?

Line 325: "If the objective is to improve the carbon cycle accuracy, the inclusion of P is advisable for its limiting role in tropical regions. From a carbon budget estimations view, we observed similar results for CN and CNP." I find this to be a very overstated conclusion. There are several studies pointing to the very uncertain representation, particularly of the P cycle in land-surface/vegetation models. Several C costs of acquiring N and P are not accounted for in these models (e.g., C allocation to mycorrhizas, enzymes, and other root exudates). These could highly impact projections of biomass growth in tropical forests (e.g., Fleischer et al., 2019; Braghiere et al., 2022). Additionally, the vegetation models poorly represent the links between N and P cycles (e.g., N cost for phosphatase enzyme production for P acquisition from leaves and soil, or a P cost for N fixation), which could also affect global projections. Moreover, there are several uncertainties about P pool dynamics in models and their fixed ratios. This is to say biomass growth in some vegetation models could be overestimated, particularly in tropical regions, especially in future scenarios. Based on your CNP vegetation/terrestrial model, could such limitations, if overcome, change your results and affect the remaining C budget, for instance? In other words, is it true the P cycle does not make a difference, or the problem is that it is loosely represented due to a lack of data? This should be discussed.

Literature

Braghiere, R.K., Fisher, J.B., Allen, K., Brzostek, E., Shi, M., Yang, X., Ricciuto, D.M., Fisher, R.A., Zhu, Q. and Phillips, R.P., 2022. Modeling global carbon costs of plant nitrogen and phosphorus acquisition. Journal of Advances in modeling earth systems, 14(8), p.e2022MS003204.

Fleischer, K., Rammig, A., De Kauwe, M.G., Walker, A.P., Domingues, T.F., Fuchslueger, L., Garcia, S., Goll, D.S., Grandis, A., Jiang, M. and Haverd, V., 2019. Amazon forest response to CO2 fertilization dependent on plant phosphorus acquisition. Nature Geoscience, 12(9), pp.736-741.

Citation: https://doi.org/10.5194/bg-2023-96-RC1
- AC1: 'Reply on RC1', Makcim Luis De Sisto Lelchitskaya, 07 Nov 2023
  
  Please find the response attached.
  
  Citation: https://doi.org/10.5194/bg-2023-96-AC1
RC2:
'Comment on bg-2023-96', Anonymous Referee #2, 11 Sep 2023
Review of “Effect of terrestrial nutrient limitation on the estimation of the remaining carbon budget” by De Sisto and MacDougall
Nutrient limitations, notably nitrogen and phosphorus, significantly regulate terrestrial carbon fixation by vegetation. Incorporating these nutrients into Earth system models is crucial for accurate climate projections. De Sisto and MacDougall assessed the remaining carbon budget using an Earth system model limited by nitrogen and phosphorus cycles, compared to a scenario without nutrient limitations. They explored various scenarios and found that nutrient limitations consistently reduced the remaining carbon budget for both 1.5°C and 2°C temperature targets compared to the model without limitations. The reduction ranged from 19% to 26%, emphasizing the importance of accounting for terrestrial nutrient limitations when estimating carbon budgets and highlighting them as a critical uncertainty in Earth system model simulations. The study is well-designed, and the results are clearly presented. I have several comments for the improvement of the manuscript.
Concerns:
Throughout the manuscript, it is not very clear how convincing those numbers presented regarding the remaining carbon budget are. It would be beneficial if the authors could show some validation after incorporating the dynamics of the N and P cycles. Additionally, providing more information about the underlying mechanisms within the model would enhance the depth of understanding of the findings. For instance, how does the model deal with carbon partitioning between growth, respiration, and storage in response to nutrient availability and other meteorological factors? Are seasonal variations in nutrient concentrations or in the C:N:P ratio in plants implemented in the model? Is there resorption after tissue senescence? And to what extent these assumptions in the model will affect the results?

In the introduction, the authors emphasized the limitation of P is crucial when modeling vegetation carbon uptake at low latitude regions (L67-69). However, in the discussion, the major impact of P limitation was concluded to be on the land use change emissions (L327-328). Is this conclusion consistent with the initial emphasis? If available, the authors could present data or analysis that demonstrates the specific regions with land use changes under P limitation.

Specific comments:
L63: The 1pctCO2 experiment was not introduced.

L80-81: What are the non-CO₂ forcings considered in the study? The impact on land use change emissions was shown in section 3.4, but the carbon cycle feedback was not fully addressed.

L85: The last paragraph in the introduction appears somewhat redundant with the first two paragraphs. Please consider consolidating or simplifying it.

L107-109: Are there distinct representations of tropical and temperate trees among the five PFTs in the model, particularly concerning their response to variations in nutrient availability and climate?

L170-171: It is not clear why the 50^th and the mean ZEC for 100 years after emissions have ceased are considered. Please clarify or add references.

L182: Please clarify what the missing forcing means.

L184-185: Is the albedo change only considered as a byproduct of land use change in this study? How about the change of albedo due to the change of LAI?

L195-196: This result has not been presented in the manuscript.

L232-234: Why the temperature exhibits a decline and subsequent increase around the 70-80^th year in Figure 3 has not been explained.

L248: There are no B2 and B3 in the Appendix. In some scenarios, C-only shows a higher temperature increase than the other two experiments (Figure B1). Could you please explain that?

L276: It was not shown in the result where and how much the land surface albedo changes due to the nutrient limitation.

L282: What does “the vegetation biomass dies” mean?

Technical corrections:
L50: Closing punctuation is missing after “…two perspectives”.

L84: “source” instead of “sources”; “accounted” instead of “account”

L145: Please add a comma after “…carbon budget estimates”.

L205: “our” instead of “out”?

L213-214: Incomplete sentence.

L222: “…is constrained…”

L224: “affect” instead of “effect”. Please also check other cases in the text.

L226: “…fluxes change…”

L227-228: “a more sensitive model”

L241: Two “budgets” in a sentence. There are also several cases elsewhere.

L261: “…clearly sensitive…”

L303: Add a comma after “…in other models”.

L307-309: Can the first two sentences be combined?

L309-310: “Mainly impacting…from land and ocean” is an incomplete sentence.

L320-321: The last sentence could be more fluid.
Citation: https://doi.org/10.5194/bg-2023-96-RC2
- AC3: 'Reply on RC2', Makcim Luis De Sisto Lelchitskaya, 07 Nov 2023
  
  Please find the response attached.
  
  Citation: https://doi.org/10.5194/bg-2023-96-AC3
RC3:
'Comment on bg-2023-96', Anonymous Referee #3, 14 Sep 2023

Effect of terrestrial nutrient limitation on the estimation of the remaining carbon budget
by De Sisto and MacDougall

General Points:

This study evaluates the impact of including nutrient limitation (in particular N and P) in estimations of the remaining carbon budget, using an Earth System Model (ESM) driven by different future Shared Socioeconomic Pathways (SSP). It tests if the land carbon sink, when including commonly neglected but relevant ecosystem processes such as the P cycle, differs from more simple implementations, such as those including only N-limitation or no limitation at all (C-only). The inclusion of P-limitation, as well as N-limitation, were expected to reduce estimations of carbon budget, due to their constraint of vegetation productivity and biomass.
The study is well described with regards to the estimations of the carbon budget, the process of vegetation limitation due to nutrients, and recent metrics of carbon accounting, such as transient climate response. The study however lacked the inclusion of references which show the effect of P in modelled ecosystem carbon responses, such as those in Dynamic Global Vegetation Models (DGVMs) (Fleischer et al., 2019; Jiang et al., 2020). The methods were clear and understandable.
My main concern was with the analysis and interpretation of results, as well as the implementation of phosphorus limitation. As far as I understood, there is no direct impact of P in photosynthesis, only N, contrary to what is described in the body of knowledge of plant physiology (Ghannoum et al., 2008; Hidaka & Kitayama, 2013; Walker et al., 2014). This may have unrealistically led the CN simulation results close to the CNP ones, and led to the conclusion that these were similar (L325). Related to this another major concern is the lack of statistical error estimations (in the case of the comparisons with the GISS temperature dataset in Figure 1), and statistical tests. While it seems clear that the CN and CNP versions were distinct from the C-only, only a formal comparison of the former would allow to reach a conclusion that CN differed or not from CNP. With such a comparison, and the effect of P in photosynthesis, it would become much clear if adding P to the model is relevant or not for the estimation of the carbon budget.

References:
Fleischer K, Rammig A, De Kauwe MG, Walker AP, Domingues TF, Fuchslueger L, Garcia S, Goll DS, Grandis A, Jiang M, et al. 2019. Amazon forest response to CO2 fertilization dependent on plant phosphorus acquisition. Nature Geoscience 12: 736–741.
Ghannoum O, Paul MJ, Ward JL, Beale MH, Corol DI, Conroy JP. 2008. The sensitivity of photosynthesis to phosphorus deficiency differs between C3 and C4 tropical grasses. Functional Plant Biology 35: 213–221.
Hidaka A, Kitayama K. 2013. Relationship between photosynthetic phosphorus-use efficiency and foliar phosphorus fractions in tropical tree species. Ecology and Evolution 3: 4872–4880.
Jiang M, Caldararu S, Zhang H, Fleischer K, Crous KY, Yang J, De Kauwe MG, Ellsworth DS, Reich PB, Tissue DT, et al. 2020. Low phosphorus supply constrains plant responses to elevated CO2: A meta-analysis. Global Change Biology 26: 5856–5873.
Walker AP, Beckerman AP, Gu L, Kattge J, Cernusak LA, Domingues TF, Scales JC, Wohlfahrt G, Wullschleger SD, Woodward FI. 2014. The relationship of leaf photosynthetic traits - Vcmax and Jmax - to leaf nitrogen, leaf phosphorus, and specific leaf area: A meta-analysis and modeling study. Ecology and Evolution 4: 3218–3235.

Citation: https://doi.org/10.5194/bg-2023-96-RC3
- AC2: 'Reply on RC3', Makcim Luis De Sisto Lelchitskaya, 07 Nov 2023
  
  Please find the response attached.
  
  Citation: https://doi.org/10.5194/bg-2023-96-AC2

Makcim L. De Sisto and Andrew H. MacDougall

Viewed

Total article views: 678 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
467	163	48	678	34	30

HTML: 467
PDF: 163
XML: 48
Total: 678
BibTeX: 34
EndNote: 30

Views and downloads (calculated since 04 Jul 2023)

Month	HTML	PDF	XML	Total
Jul 2023	136	35	4	175
Aug 2023	33	9	1	43
Sep 2023	87	37	5	129
Oct 2023	26	9	3	38
Nov 2023	39	14	10	63
Dec 2023	17	11	4	32
Jan 2024	13	6	1	20
Feb 2024	14	6	1	21
Mar 2024	27	15	2	44
Apr 2024	17	1	4	22
May 2024	20	13	7	40
Jun 2024	38	7	6	51

Cumulative views and downloads (calculated since 04 Jul 2023)

Month	HTML	PDF	XML	Total
Jul 2023	136	35	4	175
Aug 2023	33	9	1	43
Sep 2023	87	37	5	129
Oct 2023	26	9	3	38
Nov 2023	39	14	10	63
Dec 2023	17	11	4	32
Jan 2024	13	6	1	20
Feb 2024	14	6	1	21
Mar 2024	27	15	2	44
Apr 2024	17	1	4	22
May 2024	20	13	7	40
Jun 2024	38	7	6	51

Viewed (geographical distribution)

Total article views: 652 (including HTML, PDF, and XML) Thereof 652 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 29 Jun 2024

Short summary

There is uncertainty about the amount of CO₂ that can still be emitted to reach specific temperature targets. One source of uncertainty is the representation of the carbon cycle. We assessed the impact of terrestrial nitrogen and phosphorus limitation. We found a reduction in the amount of CO₂ that can still be emitted to reach temperature targets in the nutrient limited simulations. We found that nutrient limitation is an important factor to consider when estimating remaining carbon budgets.


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