the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Jul 2026
Quantifying rhizosphere priming effects on soil organic matter decomposition in situ in a subarctic ecotone using natural abundance radiocarbon
Abstract. Changes in plant-soil interactions associated with shifts in vegetation composition and climate change may have a range of effects on soil microbial activity, including increases (positive priming), inhibition (negative priming), or no net change in organic matter decomposition. The carbon-rich soils, including peats, of high latitude ecosystems, in particular, are at risk of large carbon losses linked to ongoing vegetation shifts, yet their vulnerability to priming in situ remains unresolved. Here we deploy a field-based technique, which harnesses the contemporary atmosphere as a radiocarbon (14C) ‘label’ together with a 14C-depleted (‘ancient’) peat substrate, to quantify soil organic matter (SOM) decomposition in the presence or absence of roots and rhizosphere processes in subarctic Sweden. Collars encased with different mesh sizes were placed in control and girdled (in which belowground carbon transport from the plant canopy was disrupted) mountain birch forest and willow shrub stands to test the hypothesis that the presence of ectomycorrhizal roots and extra-radical mycorrhizal mycelium increases SOM decomposition through positive priming. As expected, carbon dioxide (CO2) and dissolved organic carbon (DOC) from root ingrowth cores were significantly enriched in 14C (contemporary carbon) compared to CO2 and DOC from root exclusion cores (peat carbon), allowing partitioning of carbon mobilisation between heterotrophic (peat substrate) and recent autotrophic (plant) sources. Neither vegetation community (birch or willow), nor girdling treatment, were statistically significant as main effects, but there was a significant rhizosphere priming effect ratio of 1.36 across all groups; thus, the ancient peat-derived CO2 flux was 36 % higher in the presence of a rhizosphere than when it was absent. The lack of a significant girdling effect did not support our specific hypothesis that the presence of ectomycorrhizal roots and their associated mycelium increases SOM decomposition, but the substantial variability of modelled ancient CO2 efflux and DOC concentration during the peak growing season is consistent with the existence of ‘hot-spots’ of microbial activity. Our study provides a potential alternative to artificial substrate (e.g. glucose) additions, or 13C labelling (e.g. pulse-chase), to estimate priming in ecosystems. Furthermore, the study, undertaken in situ in the subarctic, emphasizes that increased primary productivity and associated rhizosphere processes, associated with shifts in vegetation composition and climate change, may not translate simply into increased C sequestration at whole-ecosystem scale.
- Preprint
(1308 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-3803', Erik Hobbie, 03 Aug 2026
-
RC2: 'Comment on egusphere-2026-3803', Anonymous Referee #2, 18 Aug 2026
This is a technically ambitious and genuinely novel field study. Using a ~2700-year-old, ¹⁴C-depleted peat substrate as an in situ tracer to partition rhizosphere-primed versus "ancient" CO₂ efflux is a very interesting way for field quantification of rhizosphere priming effects, complementing (rather than duplicating) the lab-based ¹³C-labelling. The pairing with an independent girdling manipulation is a strong design feature, and the treatment of atmospheric contamination (Keeling-plot correction and sensitivity analysis) are commendable examples of rigorous, transparent radiocarbon methodology.
Main comments
1- I agree with Reviewer 1's assessment and specific comments. The figures do need revision for colour-independent interpretability, the terminology issued must be double checked. I would add that the manuscript is internally inconsistent on this last point, and that the "enriched/depleted" imprecision recurs beyond the lines Reviewer 1 flagged.
2- The girdling treatment had likely waned by the time of sampling. Root biomass showed no significant girdling effect by 2021, and the authors themselves attribute this to compensatory understorey rhizosphere development over the (unplanned) fourth growing season. This means Hypothesis 2 was not actually tested under the conditions assumed by the original design. The "no significant girdling effect" result cannot be interpreted as evidence against a role for canopy ectomycorrhizal hosts, only as evidence that the girdling manipulation itself had become confounded by regrowth. This is discussed in the Discussion but should be reflected more explicitly in the abstract/conclusion, where the current phrasing risks being read as a clean negative result rather than an inconclusive one.
3- The peat-derived (Δpeat) end-member is a single, once-measured laboratory value applied uniformly to ~40+ field cores across 4 years and heterogeneous microsites, with no propagation of its uncertainty into the mixing-model outputs, and no empirical test of whether it remained representative after 4 years of in situ decomposition. Given that the authors already ran a valuable sensitivity analysis for the Δair end-member, a parallel sensitivity analysis varying Δpeat within a plausible range would substantially strengthen confidence in the robustness of the RPE estimate.
4- The root-severing validation (Appendix A) is a reasonable proof-of-concept but should be explicitly caveated as an approximation, not a direct validation of the subarctic peat-core system.
5- The mesh-based ingrowth/exclusion design cannot distinguish whether priming arises from ectomycorrhizal enzymatic SOM mining, rhizodeposition-stimulated saprotrophic activity, or root-associated fauna. Yet the hypotheses are framed specifically around ectomycorrhizal roots and extra-radical mycelium. Given that the willow understorey also contains an ECM species (Betula nana), the canopy vs. understorey ECM contrast is not clean. This should be acknowledged as a limit on mechanistic attribution, not just a footnote about uncontrolled species composition.
6- All CO₂/DOC/¹⁴C measurements come from one week in July 2021. The discussion's invocation of "hot moments" (temporal heterogeneity) is not directly testable with this design. Only spatial heterogeneity ("hotspots") is empirically demonstrated. I'd recommend softening this language.
Specific comments
L. 106–110: Six birch pairs and five willow pairs are stated here, but Fig. 2/Table B1 show willow girdled n = 4.
L. 130–136 / (L. 179–186): Please clarify whether the bulk peat and the incubation-derived Δpeat sample were homogenized/composited to ensure the same ¹⁴C signature applies across all field cores, and report whether any replicate ¹⁴C measurements of the peat batch were made. As it stands, Δpeat is a single analytical value (with only counting-statistics uncertainty, ±0.36) used as a fixed parameter in Eq. 3 across ~40 cores over 4 years; given the acknowledged possibility of shifting lability/age of the decomposing peat pool (rows 437–441), I recommend a sensitivity analysis analogous to the one already performed for Δair, testing plausible bounds on Δpeat.
L. 152–156: Please note explicitly that this is a single sampling date/growing season, which limits the ability to draw conclusions about temporal variability as opposed to purely spatial variability.
L. 168–171: The CO₂ build-up/incubation duration differed substantially between ingrowth (2–4 h) and exclusion cores (up to 12 h). Please discuss whether this differential incubation time could itself introduce a systematic artifact (e.g., progressive substrate-pool switching, changing headspace conditions) that covaries with treatment, independent of the biological priming effect being tested.
L. 179–186: Please state the temperature and moisture conditions during the in vitro peat incubation relative to in situ field conditions at Abisko during the July sampling window, since basal respiration rate and community composition are temperature/moisture sensitive and could affect how representative this laboratory Δpeat value is of actual field decomposition.
L. 336: This is a strong claim. I suggest tempering it in light of the limitations above (small n, single time-point, confounded girdling treatment by year 4, single-measurement peat end-member).
L. 385–415: This paragraph does a good job acknowledging the girdling-effect waning issue. An equivalent, explicit caveat could be added to the abstract/conclusion, since as currently phrased these sections could be read as reporting a clean negative result for the girdling hypothesis rather than an inconclusive/confounded one.
Statistics: A large number of fixed-effect significance tests are reported across multiple response variables (Tables B1–B4) without correction for multiple comparisons. I don't think this changes the paper's main conclusions, but a sentence acknowledging this (and perhaps highlighting which effects remain robust under a more conservative threshold) would strengthen the statistical rigour of the paper.
Citation: https://doi.org/10.5194/egusphere-2026-3803-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 89 | 43 | 12 | 144 | 17 | 11 |
- HTML: 89
- PDF: 43
- XML: 12
- Total: 144
- BibTeX: 17
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
In this study, the authors used 2700-year-old peat cores to study rhizosphere priming in the subarctic in willow and birch communities. The study was carefully done, although like many radiocarbon studies, sample sizess are small. The use of fine-mesh cores to exclude mycorrhizal fungi allowed the contribution of new photosynthate versus old peat to be roughly assessed. This approach was complemented by concurrent girdling of source shrubs/trees (willow/birch) in the experiment. These mesh cores allowed rhizosphere priming effects to be assessed, with an overall effect of 1.36, or 36% more CO2 flux from cores with the rhizosphere connections present than with those connections not present. It was surprising that girdling of willow or birch had very little effect on the measured CO2 fluxes.
The paper would be improved particularly by focusing on the figures. Figure 1, the schematic, was very good, others could be easily improved. Figure information should ideally not be lost if printed in black and white, as I did. For example, in Figure 2, 2a and 2b should be labeled clearly as birch, 2c and 2d labeled clearly as willow. Similarly, lines in Figures 2b and 2d indicating 14C of atmospheric CO2 and peat-derived CO2 should not require color. Just label the lines. Similar improvements could be done to Figure 3. For Figure 4, perhaps the left four columns could be labeled birch and the right 4 labeled willow.
General. Equal, etc.s signs (=, <, >) should have adjoining spaces, these are mathematical sentences.