Evaluation of simulated N cycling using observations from a <sup>15</sup>N tracer experiment in a mixed deciduous forest

Thum, Tea; Goodale, Christine L.; Yu, Lin; Nabel, Julia; Zaehle, Sönke

doi:10.5194/egusphere-2025-5876

Preprints

https://doi.org/10.5194/egusphere-2025-5876

Preprints

04 Dec 2025

| 04 Dec 2025

Evaluation of simulated N cycling using observations from a ¹⁵N tracer experiment in a mixed deciduous forest

Tea Thum, Christine L. Goodale, Lin Yu, Julia Nabel, and Sönke Zaehle

Abstract. Nitrogen availability constrains the terrestrial carbon uptake and storage, yet large uncertainties remain in the magnitude of the effect, because the interactions of the carbon and nitrogen (N) dynamics are challenging to observe in undisturbed ecosystems at relevant timescales. Long-term experiments with ¹⁵N tracer applications allow study of the nitrogen cycle in a fairly undisturbed manner, and they are therefore a valuable data source to test the biogeochemical dynamics simulated by terrestrial biosphere models. In this study we applied the model QUINCY (QUantifying Interactions between Terrestrial Nutrient CYcles and the climate system), which includes an explicit representation of terrestrial ¹⁵N fluxes and pools. We used observations from a long-term (10-year) ¹⁵N tracer experiment in a temperate deciduous forest to evaluate the nitrogen dynamics simulated by QUINCY. Recovery in soil N dominated overall ecosystem ¹⁵N recovery in both observations and simulations over the long-term. The observed gradual movement of the ¹⁵N tracer to lower soil layers was also captured by the model. However, in the short-term modeled uptake and losses of ¹⁵N into leaves and fine roots were too fast, and recovery in litter and surface soil was too slow, indicating that the model likely overestimates plant competitiveness for newly added N relative to soil microbes. Downward vertical transport of ¹⁵N tracer in the soil was slow compared to measurements, which may be indicative either of too low bioturbation or vertical transport via leaching. Overall, the QUINCY model results showed good agreement with the observations making it a valuable tool to study long term nitrogen dynamics. Running the model for an extended period for example indicated that the ecosystem retained a very large share of the added ¹⁵N tracer (>90 %), and that this retention persisted at multi-decadal timescales. This study shows that explicit inclusion of isotopic tracers allows for a more profound evaluation of carbon-nitrogen turnover and dynamics and thereby can contribute to reduce uncertainties in modelling nitrogen flow and constrains in terrestrial ecosystems.

Received: 27 Nov 2025 – Discussion started: 04 Dec 2025

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 paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 547 KB)

Supplement (57 KB)

Download & links

Tea Thum, Christine L. Goodale, Lin Yu, Julia Nabel, and Sönke Zaehle

Status: closed

RC1:
'Comment on egusphere-2025-5876', Anonymous Referee #1, 04 Feb 2026
General Comments:
This is a thorough and well-conceived paper describing efforts to use experimental observations to validate a terrestrial biosphere model, with the goal of evaluating how effectively this model describes ecosystem retention of excess nitrogen. The authors use results from a ¹⁵N tracer experiment, where plant and soil N retention was measured three times over ten years, to evaluate the QUINCY model’s ability to describe the movement of N through a temperate deciduous forest ecosystem.
The authors are quite thorough in their discussion of the model and how it relates to field observations over time, and do an effective job at explaining where mismatches between simulated and observed N pools occur. The science presented here is sound, and I do not see any critical issues in the methods or interpretation of results in this paper. The work described here is a solid contribution to the literature and should be of interest to BG readers.
My primary critiques are with the contextualization of the study and the results. While the bulk of the paper is spent discussing the model behavior in relation to field observations, there needs to be more attention given to the implications of this work, particularly in the Discussion.
Specific Comments by section:
Introduction
Lines 29-43: Authors should acknowledge long-term decreasing trend in N deposition in certain temperate regions due to air pollution legislation, particularly in Europe and North America. This is particularly relevant given the location of the study in the Northeast U.S. See Mason et al. (2022; https://doi.org/10.1126/science.abh3767) and related literature, including Craine et al. (2018; https://doi.org/10.1038/s41559-018-0694-0) and Groffman et al. (2018; https://doi.org/1007/s10533-018-0445-y).

A little more background on the QUINCY model is needed in the introduction. We need to understand more about why this model is chosen. Is it the only model that includes an explicit ¹⁵N cycle? If not, is there anything of note that makes this model most suited to simulating results from this experiment?

Methods
Line 93-94: What is the background N deposition at the site? Generally, what are the trends in that deposition? Is it increasing or decreasing, and how does it compare to places with high N deposition?

Results
Table 1: Color coding to quickly show under or over estimation would be helpful

Figure 1: would benefit from larger panels. The figure is hard to read as it is.

Table 2: Same as Table 1, use color coding for under and over estimation

Discussion
Section 4.1 is very dense, I recommend splitting with subheadings to make it more digestible.

Line 297-298: Do you have a sense for why the model underestimated these C and N pools? If so, that should be stated.

Lines 316-318: There should be some further explanation for this, as 2mm is typically the standard cutoff for fine roots. Do the authors believe that this impacted results at all? Either way, the effect of this discrepancy, or lack thereof, should be specified.

Lines 401-408: This presents the question of why you did not use the JSM rather than the basic soil model in QUINCY. The authors should provide a justification for not choosing JSM (benefits of simpler parameters, computational power, etc.)

Conclusion
This should be tightened up, as it reads too much like a restating of the discussion.

The paper ends abruptly and would benefit from a sentence or two summing up the primary findings, stating the implications of those findings, and suggesting future directions for this work (similar to what was stated in the abstract).

Technical Corrections:
Line 213: Typo, “The observations showed a smaller enrichment peak as the simulations”, replace “as” with “than”

Line 345 & 353: Typo, should be ¹⁵N

Line 346-348: Sentence is a bit confusing and contains a typo on line 347
Citation: https://doi.org/10.5194/egusphere-2025-5876-RC1
- AC1: 'Reply on RC1', Tea Thum, 13 Mar 2026
  
  We thank the reviewer for these insightful comments and have added point to point reply in the attachment.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5876-AC1
RC2:
'Comment on egusphere-2025-5876', Anonymous Referee #2, 13 Feb 2026

This manuscript tracks nitrogen dynamics within a temperate deciduous broadleaf forest by running a 15N tracer addition experiment both at a site in New York and a paired experiment in the QUINCY land model. The observational data are used to evaluate the nitrogen cycling processes in QUINCY and include 15N measurements across soil and plant pools. The observations span approximately 10 years and the model simulations match this time frame while also including a simulation at 30 years following the end of the observational data.
Overall the model does a good job at tracking the observations and where there are deviations the authors do a good job at explaining the why model and observations may deviate and how this information can be used to improve the QUINCY model. The study is well conducted, and the manuscript is well written. The topic is also to be of high interest to readers and is a good fit for Biogeosciences. In general, my comments are minor.
L1-2: First sentence seems a little circular (i.e., climate change influences… new conditions due to changing climate). Rewrite.
L51: Can you provide a citation or more detailed explanation of optimal N fertilization for forest management?
L79: Change “will” to “does” in the question, as the question includes the present.
L110: Although this paper used deciduous temperate broadleaf as a PFT, from this sentence it is unclear if a site (or for simulations at larger spatial scales, a grid cell) can be represented by more than one PFT (i.e., a mix of PFTs).
L134-135: BNF is mentioned but my understanding is that it was not simulated for this study. Does the model allow BNF to be switched on and off for a PFT or is it always happening or does it occur in some PFTs but not others? Also, is there asymbiotic, free-living BNF?
L174-175: Given that N deposition is prescribed and not affected by the processes in the ecosystem, perhaps this sentence should be moved to the Methods.
L218-220: This sentence is unclear. It states that fine root d15N is higher in the top layer for (0-10 cm), in part because there is a lof of fine root biomass at 10-20 cm. That does not make sense to me as it suggests the high root biomass at 10-20 cm should decrease d15N in the top layer.
L254: Add “soil” after “simulated”.
L282-283: In addition to the simulations, the observations also point to a fairly closed nitrogen cycle (excluding the initial loss in recovery following immediate application but looking at the differences in Figure 4a between 2008, 2012, 2017).
L334-336: I think it would be clearer to say “This influence was shown by isolating the fast pools…”.
L381-382: This sentence is unclear.
L397-400: It would be helpful to comment on whether trees at the site form ectomycorrhizal or arbuscular mycorrhizal symbioses.
L409 and 534-537: The authors names should be written for this reference (instead of just the abbreviation).
Can it be inferred that N limitation is stronger in the simulations than the observations? Figure 3c shows higher soil N in the simulation and figure 1e shows higher aboveground pools and lower belowground pools in the model (i.e., greater relative investment in aboveground tissues in the model).
I think the manuscript would benefit from a greater discussion of the implications of the study, particularly in the Conclusions section.
Sentences in the main text that introduce the figures or tables are unnecessary (e.g., L206, L253, L264, L287-288). However, upon deleting them, some sentences will need to explicitly mention figures (e.g., in parentheses).
Sometimes the ¹⁵ in ¹⁵N is not superscript.
Figure 1c shows some 15N in coarse roots in the model simulations. This seems inconsistent with Figure 4b which shows no recovery in coarse roots in model simulations.
Figure 1 caption: There is no panel f. As for the description of panel d, should that be a combination of sapwood and heartwood?
Figure 3c: Is this total soil N (inorganic plus organic)?

Citation: https://doi.org/10.5194/egusphere-2025-5876-RC2
- AC2: 'Reply on RC2', Tea Thum, 13 Mar 2026
  
  We thank the reviewer for these insightful comments and have added a point to point reply in the attachment.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5876-AC2

Status: closed

RC1:
'Comment on egusphere-2025-5876', Anonymous Referee #1, 04 Feb 2026
General Comments:
This is a thorough and well-conceived paper describing efforts to use experimental observations to validate a terrestrial biosphere model, with the goal of evaluating how effectively this model describes ecosystem retention of excess nitrogen. The authors use results from a ¹⁵N tracer experiment, where plant and soil N retention was measured three times over ten years, to evaluate the QUINCY model’s ability to describe the movement of N through a temperate deciduous forest ecosystem.
The authors are quite thorough in their discussion of the model and how it relates to field observations over time, and do an effective job at explaining where mismatches between simulated and observed N pools occur. The science presented here is sound, and I do not see any critical issues in the methods or interpretation of results in this paper. The work described here is a solid contribution to the literature and should be of interest to BG readers.
My primary critiques are with the contextualization of the study and the results. While the bulk of the paper is spent discussing the model behavior in relation to field observations, there needs to be more attention given to the implications of this work, particularly in the Discussion.
Specific Comments by section:
Introduction
Lines 29-43: Authors should acknowledge long-term decreasing trend in N deposition in certain temperate regions due to air pollution legislation, particularly in Europe and North America. This is particularly relevant given the location of the study in the Northeast U.S. See Mason et al. (2022; https://doi.org/10.1126/science.abh3767) and related literature, including Craine et al. (2018; https://doi.org/10.1038/s41559-018-0694-0) and Groffman et al. (2018; https://doi.org/1007/s10533-018-0445-y).

A little more background on the QUINCY model is needed in the introduction. We need to understand more about why this model is chosen. Is it the only model that includes an explicit ¹⁵N cycle? If not, is there anything of note that makes this model most suited to simulating results from this experiment?

Methods
Line 93-94: What is the background N deposition at the site? Generally, what are the trends in that deposition? Is it increasing or decreasing, and how does it compare to places with high N deposition?

Results
Table 1: Color coding to quickly show under or over estimation would be helpful

Figure 1: would benefit from larger panels. The figure is hard to read as it is.

Table 2: Same as Table 1, use color coding for under and over estimation

Discussion
Section 4.1 is very dense, I recommend splitting with subheadings to make it more digestible.

Line 297-298: Do you have a sense for why the model underestimated these C and N pools? If so, that should be stated.

Lines 316-318: There should be some further explanation for this, as 2mm is typically the standard cutoff for fine roots. Do the authors believe that this impacted results at all? Either way, the effect of this discrepancy, or lack thereof, should be specified.

Lines 401-408: This presents the question of why you did not use the JSM rather than the basic soil model in QUINCY. The authors should provide a justification for not choosing JSM (benefits of simpler parameters, computational power, etc.)

Conclusion
This should be tightened up, as it reads too much like a restating of the discussion.

The paper ends abruptly and would benefit from a sentence or two summing up the primary findings, stating the implications of those findings, and suggesting future directions for this work (similar to what was stated in the abstract).

Technical Corrections:
Line 213: Typo, “The observations showed a smaller enrichment peak as the simulations”, replace “as” with “than”

Line 345 & 353: Typo, should be ¹⁵N

Line 346-348: Sentence is a bit confusing and contains a typo on line 347
Citation: https://doi.org/10.5194/egusphere-2025-5876-RC1
- AC1: 'Reply on RC1', Tea Thum, 13 Mar 2026
  
  We thank the reviewer for these insightful comments and have added point to point reply in the attachment.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5876-AC1
RC2:
'Comment on egusphere-2025-5876', Anonymous Referee #2, 13 Feb 2026

This manuscript tracks nitrogen dynamics within a temperate deciduous broadleaf forest by running a 15N tracer addition experiment both at a site in New York and a paired experiment in the QUINCY land model. The observational data are used to evaluate the nitrogen cycling processes in QUINCY and include 15N measurements across soil and plant pools. The observations span approximately 10 years and the model simulations match this time frame while also including a simulation at 30 years following the end of the observational data.
Overall the model does a good job at tracking the observations and where there are deviations the authors do a good job at explaining the why model and observations may deviate and how this information can be used to improve the QUINCY model. The study is well conducted, and the manuscript is well written. The topic is also to be of high interest to readers and is a good fit for Biogeosciences. In general, my comments are minor.
L1-2: First sentence seems a little circular (i.e., climate change influences… new conditions due to changing climate). Rewrite.
L51: Can you provide a citation or more detailed explanation of optimal N fertilization for forest management?
L79: Change “will” to “does” in the question, as the question includes the present.
L110: Although this paper used deciduous temperate broadleaf as a PFT, from this sentence it is unclear if a site (or for simulations at larger spatial scales, a grid cell) can be represented by more than one PFT (i.e., a mix of PFTs).
L134-135: BNF is mentioned but my understanding is that it was not simulated for this study. Does the model allow BNF to be switched on and off for a PFT or is it always happening or does it occur in some PFTs but not others? Also, is there asymbiotic, free-living BNF?
L174-175: Given that N deposition is prescribed and not affected by the processes in the ecosystem, perhaps this sentence should be moved to the Methods.
L218-220: This sentence is unclear. It states that fine root d15N is higher in the top layer for (0-10 cm), in part because there is a lof of fine root biomass at 10-20 cm. That does not make sense to me as it suggests the high root biomass at 10-20 cm should decrease d15N in the top layer.
L254: Add “soil” after “simulated”.
L282-283: In addition to the simulations, the observations also point to a fairly closed nitrogen cycle (excluding the initial loss in recovery following immediate application but looking at the differences in Figure 4a between 2008, 2012, 2017).
L334-336: I think it would be clearer to say “This influence was shown by isolating the fast pools…”.
L381-382: This sentence is unclear.
L397-400: It would be helpful to comment on whether trees at the site form ectomycorrhizal or arbuscular mycorrhizal symbioses.
L409 and 534-537: The authors names should be written for this reference (instead of just the abbreviation).
Can it be inferred that N limitation is stronger in the simulations than the observations? Figure 3c shows higher soil N in the simulation and figure 1e shows higher aboveground pools and lower belowground pools in the model (i.e., greater relative investment in aboveground tissues in the model).
I think the manuscript would benefit from a greater discussion of the implications of the study, particularly in the Conclusions section.
Sentences in the main text that introduce the figures or tables are unnecessary (e.g., L206, L253, L264, L287-288). However, upon deleting them, some sentences will need to explicitly mention figures (e.g., in parentheses).
Sometimes the ¹⁵ in ¹⁵N is not superscript.
Figure 1c shows some 15N in coarse roots in the model simulations. This seems inconsistent with Figure 4b which shows no recovery in coarse roots in model simulations.
Figure 1 caption: There is no panel f. As for the description of panel d, should that be a combination of sapwood and heartwood?
Figure 3c: Is this total soil N (inorganic plus organic)?

Citation: https://doi.org/10.5194/egusphere-2025-5876-RC2
- AC2: 'Reply on RC2', Tea Thum, 13 Mar 2026
  
  We thank the reviewer for these insightful comments and have added a point to point reply in the attachment.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5876-AC2

Tea Thum, Christine L. Goodale, Lin Yu, Julia Nabel, and Sönke Zaehle

Supplement

https://doi.org/10.5194/egusphere-2025-5876-supplement

Tea Thum, Christine L. Goodale, Lin Yu, Julia Nabel, and Sönke Zaehle

Viewed

Total article views: 2,165 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
1,358	630	177	2,165	289	142	131

HTML: 1,358
PDF: 630
XML: 177
Total: 2,165
Supplement: 289
BibTeX: 142
EndNote: 131

Views and downloads (calculated since 04 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	449	179	65	693
Jan 2026	129	140	40	309
Feb 2026	302	100	36	438
Mar 2026	345	131	19	495
Apr 2026	68	30	3	101
May 2026	26	21	2	49
Jun 2026	11	4	0	15
Jul 2026	12	19	1	32
Aug 2026	10	4	10	24
Sep 2026	6	2	1	9

Cumulative views and downloads (calculated since 04 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	449	179	65	693
Jan 2026	129	140	40	309
Feb 2026	302	100	36	438
Mar 2026	345	131	19	495
Apr 2026	68	30	3	101
May 2026	26	21	2	49
Jun 2026	11	4	0	15
Jul 2026	12	19	1	32
Aug 2026	10	4	10	24
Sep 2026	6	2	1	9

Viewed (geographical distribution)

Total article views: 2,137 (including HTML, PDF, and XML) Thereof 2,137 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 04 Sep 2026

Short summary

Nitrogen and carbon cycle have manifold interactions and to be able predict carbon sequestration of terrestrial vegetation, ability to model nitrogen cycle properly is important. ¹⁵N tracer experiments can be helpful, as they track the fate where added nitrogen will end up in the ecosystem. We simulated one such experiment with a model containing ¹⁵N cycle and found that comparison with data helped to evaluate the model in different aspects.


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

Evaluation of simulated N cycling using observations from a 15N tracer experiment in a mixed deciduous forest

Supplement

Viewed

Viewed (geographical distribution)

Evaluation of simulated N cycling using observations from a ¹⁵N tracer experiment in a mixed deciduous forest