the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Nov 2025
CO2 emissions from dredged sediment as a function of moisture, temperature, and oxygen
Abstract. Estuaries represent a crucial compartment in the global carbon cycle, with high rates of organic matter formation, burial and degradation. Sedimentary processes control the balance between long-term burial of carbon and CO2 and CH4 emissions upon OM degradation, for which estuaries are a global hotspot. The profound and globally intensifying perturbation of estuarine sediment by anthropogenic activities such as harbor dredging has a far-reaching but poorly understood impact on sedimentary carbon cycling processes in estuaries and by extension potentially on global carbon budgets. Hence, understanding carbon emissions from dredged sediments under varying environmental conditions is critical for assessing their environmental impact and informing large-scale sediment reuse strategies. This study investigates the effects of moisture, temperature, and oxygen availability on CO2 emission rates from dredged sediments collected from the Port of Rotterdam, the largest port in Europe. Results are compared with soil CO2 emissions from a global database of nearly 400 laboratory incubations under different conditions. Our sediment incubation showed that CO₂ emissions increased 1.5–8.1 times with higher moisture levels, 3.8–6.0 times with elevated temperatures, and 4.5–6.4 times with oxygen exposure. Applying machine-learning tools (XGBoost) to a global database of soils and sediment incubations suggested that moisture and temperature responses observed in our experiment are widespread in both soils and sediments. However, functions that described these responses differed significantly from those used in global biogeochemical carbon-cycle models, indicating a need to revisit these functions. Oxygen displayed a relatively stronger effect in sediments, likely due to better preservation of labile organic matter (OM) in anoxic conditions and its rapid oxidation upon re-exposure to oxygen. A model incorporating organic matter with different degradation rates showed that while labile OM fueled high initial rates of CO2 emission, more recalcitrant OM was a much more abundant OM pool (> 80 %) that dominated cumulative CO2 emissions on longer timescales. Overall, our experiment and meta-analysis on a global soil dataset suggest the importance of environmental controls on carbon emissions and that dredged sediments are an organic-rich, potent source of CO2 upon oxidation after dredging, which should be considered in sediment management and reuse practices.
- Preprint
(1760 KB) - Metadata XML
-
Supplement
(1911 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-5709', Nathan McTigue, 20 Dec 2025
-
RC2: 'Comment on egusphere-2025-5709', Anonymous Referee #2, 09 Feb 2026
The authors present the results of an experiment in which they incubated various samples of dredged sediments from the Port of Rotterdam for approximately one month under different temperatures and water contents, as well as under oxic and anoxic conditions. They supplemented these results with an evaluation of data from a global incubation database concerning the impact of water content, temperature, and oxygen on the formation of microbial CO₂ in soils from various climatic regions. This evaluation was carried out using a predictive model trained with log-transformed CO₂ respiration rates.
The manuscript deals with the undoubtedly important topic of the influence of soil temperature, oxygen content and water content on microbial CO₂ formation. However, it attempts to bring together two very different datasets that are only remotely related. The introduction discusses the importance of estuaries and dredged sediments. It is only in the final paragraph that it is mentioned that an incubation database was also evaluated. However, the majority of the manuscript then focuses on the results of the database, to which the results of the experiment were added. In this respect, the title does not accurately reflect the content of the manuscript. This approach is problematic because the database mainly consists of results from incubations of terrestrial soils, which are only comparable with harbour sediments to a limited extent.
I am also concerned about the evaluation of the global database by a machine learning tool. The results concerning the influence of water content, temperature, and the presence of oxygen are not always comprehensible (see specific comments) and are partly biased by the distribution of soil samples in the database, which mainly originate from temperate environments. Additionally, log-transforming respiration rates reduces the difference between extreme values, resulting in a smaller effect in the predictive model than in the observations. While the authors report good predictive performance of the model (L262), they also state that the model's predictions deviate substantially from the observations (L338 ff.). I am therefore unsure what additional information we obtain from evaluating the data using the predictive model.
Furthermore, I have reservations about the method used to calculate the amount of CO2 in the bottles. At high pH, a significant proportion of inorganic carbon dissolves in the water. Neglecting this results in a significant underestimation of CO₂ respiration rates.
Finally, I strongly recommend avoiding the term 'emission' when meaning 'production', since 'emission' applies to in situ fluxes. All the data in this manuscript are only loosely related to in situ fluxes; only potential production is presented.
Specific comments:
L13: I doubt that harbour dredging has an impact on the global carbon budget. This claim should be substantiated with data, for example by comparing riverine sediment transport into estuaries with estimates of dredged sediments, or by comparing potential CO₂ fluxes from dredged sediments with CO₂ fluxes from global soils.
L14 ff: While understanding the regulation of carbon emissions from dredged sediments is certainly important, this manuscript is not about CO₂ emissions under in situ conditions from dredged sediments, but rather about potential CO₂ production in laboratory incubations. Throughout the manuscript, I strongly suggest differentiating between in situ CO₂ emissions and potential CO₂ production to prevent confusion and unsubstantiated conclusions.
L23: 'Labile' is not the right term in this context, since it is the conditions (oxic vs. anoxic) that determine whether a given C pool can be decomposed. Clearly, the organic matter (OM) in question is not labile under anoxic conditions.
L26ff: It would be better to present the novel aspects of the study or provide an outlook at the end of the abstract rather than ending with well-known facts.
38: Again, the amount of CO₂ released from dredged sediments is most likely not affecting the global carbon budget. To support their claim, the authors should present some data.
L66f: Arrhenius kinetics predict an exponential increase in decomposition rates with temperature, yet in studies on microbial carbon decomposition, this model is only applied to temperatures below the ‘optimum’ temperature, where maximum rates are observed. I am not aware of any study that assumes the absence of an 'optimum' temperature for microbial activity.
L72f: This statement is unclear to me. Why should the effects of redox state, OM and microbial communities differ in dredged sediment compared to 'native soils'?
L80ff: Please differentiate between 'emission' and 'production'.
L98ff: Please explain how the porosity of the dried samples was determined.
L101: The high concentration of H₂ in the anoxic incubations will cause an enrichment of H₂ -consuming anoxic microbes. Gas production rates under such high substrate conditions provide no information about rates under in situ conditions.
L107: An incubation period of only 37 days provides little information about in situ gas production. Preparing the sediment (freezing to -20°C, freeze-drying, grinding and rewetting) disturbs the system so much that initial microbial activities are heavily biased. Furthermore, calculating rates from only two data points introduces high uncertainty. Finally, the dissolved inorganic carbon was not considered when determining the OM decomposition rate. At high pH values (>7), most of the total amount of inorganic carbon produced during OM decomposition remains in the water as DIC rather than in the air as CO₂. This may introduce substantial errors in calculating OM decomposition rates at different water contents. Therefore, the pH value of the sediments should be reported, and the DIC in the water must be considered.
L131: What is the 'average carbon emission rate throughout the incubation' and why is a model needed for this if carbon production rates were measured?
L176ff: Fitting a two-pool model to incubations as short as seven days, or even 37 days as presented in this manuscript, will not provide any meaningful information since the incubation time is far too short. The authors could test for this bias using longer studies from the database. In general, a two-pool model will produce substantially higher decomposition rate constants when a shorter incubation period is considered, since rates are highest at the beginning of the incubation period.
L202: What do the authors mean by 'adaptation towards high water content'? Do they mean low oxygen concentration? It was not explained how the amount of water was determined to give 100% water-filled pore space. Perhaps the sediment was not fully saturated.
L203f: The desorption of OM from mineral surfaces and its diffusion are affected by moisture content. But what does this have to do with aggregates? Does water not affect the diffusion and desorption of substrates from non-aggregated particles?
L214: It is not clear how DNA extraction should determine the reason for the absence of a difference in CO₂ production rates at 20 and 30°C. A higher resolution of temperatures and a larger temperature range might help.
L240: Absolute rates of CO₂ production depend heavily on incubation experiment duration, as rates decrease significantly over time. Therefore, a simple comparison of rates from incubation experiments lasting between seven and 1,000 days does not provide meaningful information. Since the incubation experiment in this study was very short, it can be expected that the rates are in the higher range.
L267 ff: I cannot see a decline in activity above 80% WFPS. For forests, there is a clear increase in activity up to 80%, with no data above this water content. For sediments and wetlands, activity increases up to 100%. I doubt that failing to consider the different environmental conditions in different types of soil and sediments will provide valuable information about the general response of microbial activity to soil water content.
L279ff: The authors seem to assume that microbial activity exhibits a universal temperature response with a global optimum between 20 °C and 35 °C, but this is a misconception. The temperature optimum for microbial activity, such as aerobic respiration, depends heavily on the composition of the microbial community, which is in turn affected by environmental temperature and its fluctuations. The temperature minimum, optimum and maximum for microbial respiration in tundra soils are substantially lower than in tropical soils. Therefore, the highest activities may be found below 20 °C or above 40 °C, depending on the climate zone from which the soil samples originate. The results presented in Fig. 3c appear to be heavily biased by the dominance of soil samples from temperate regions.
L295 ff: The effect of oxygen on respiration rates is extraordinarily low; the ratio between oxic and anoxic decomposition is in experiments often found at or above 3. Furthermore, I would expect a greater response to oxygen availability in well-drained oxic soils, such as forest soils, since they are unlikely to harbour a high proportion of microbes adapted to anoxic conditions, as found in sediments and wetlands, which are oxic at the surface and anoxic in deeper layers. However, the opposite is true.
L432 ff: These two paragraphs are somewhat unrelated to all the preceding topics.
L456f: This is correct, but in the Discussion, the authors write that rates decreased above 80% WFPS. Please be consistent and discuss why your findings show the highest rates at water saturation, which contrasts with previous findings.
Citation: https://doi.org/10.5194/egusphere-2025-5709-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 260 | 115 | 22 | 397 | 48 | 16 | 16 |
- HTML: 260
- PDF: 115
- XML: 22
- Total: 397
- Supplement: 48
- BibTeX: 16
- EndNote: 16
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript explores factors that affect the carbon dioxide emissions of dredge spoils. There are compelling reasons, including the sheer volume of dredge spoils excavated globally, to pursue this question in terms of global carbon cycling and management strategies. The authors performed an incubation of sediment collected from dredge spoils at the Port of Rotterdam, a location where dredging is a frequent undertaking. They proceeded to compare their results with a meta-analysis of almost 400 incubations, which were also used for machine learning to determine the importance of and interaction between oxygen, moisture, and temperature. Briefly, the authors report that temperature does have an important role, but may taper out between 20 and 35 degrees C where an enzymatic optimum is reached. Oxygen played a significant role in increasing carbon emissions. Moisture was also positively correlated with decomposition where wetter sediment yielded higher respiration rates. Together, this information represents a novel perspective in considering dredge spoils as a source of carbon dioxide to the atmosphere. Moreover, their approach provides important factors for other researchers to consider when preparing similar incubations.
Overall, this manuscript is written clearly. The methods are valid and reproducible. The results are not over-interpreted yet demonstrate their importance. The figures are sufficient, but I recommend including additional information in the Supplemental to list the studies used for meta-analysis.
My most substantial comment is to elaborate on the two-pool modeling approach: it is unclear how the proportion of the slow pool and fast pool were attributed. As discussed later in the manuscript, dredge spoils oftentimes consist of older, refractory material and contain only slow-pool OM. How is the portion of each pool determined? If dredge spoils usually constitute the slow-pool, is there any downside to continue using the two-pool model?
L58: The sentence beginning “This important…” is missing a word/verb. Perhaps “This is important globally because perturbations of…”
L93: “refrigerator” instead of “fridge”.
L116: I encourage the authors to tabulate the sources of data for meta-analysis. Fig S2 is helpful, but it would be helpful to see all the publications used.
L154: bring Sierra et al. Outside of parentheses and only cite the year: Following Sierra et al. (2017), the effects…
L160: I agree with the two-pool modeling approach. Could you elaborate more about the initial fractions of the fast pool and slow pool. What if dredge spoils are old sediments and contain mostly the slow pool? This is discussed further on L380-381.
L273: meta-analysis instead of meta-study?
L284: can you elaborate more on the mechanism driving this? Is it that microbes are not adapted to temperatures higher than this since they rarely experience it? Do enzymes stop working at these temperatures? It is surprising since I have read of (and conducted myself) experiments that use 20 & 30 or 15 & 25 degrees C for incubation temperatures and measure Q10 of ~2.
L309-317: Interesting about sand content. Could the machine learning be sensitive to this since higher sand content might possess lower carbon content which could possibly be more refractory organic matter, even when standardized to C? In other words, higher sand is just a proxy for low concentrations of refractory OM that yield low decomposition rates?
L438-439: Wait, I thought anaerobic conditions *decreased* GHG emissions?