We sincerely thank the reviewer for his/her careful reading and insightful comments on our manuscript. The constructive feedback has greatly improved the clarity and rigor of this work. We have addressed all comments in detail and revised the manuscript accordingly. Below, we provide point-by-point responses to the comments. A rebuttal with screenshots of the revised text in the manuscript is included in the attached file.

MAJOR COMMENTS

[1] I feel that a clarity is needed for how much published data were taken from Lin et al. (2025a, 2025b), as the authors refer often these publications apart from some other sources. In addition, what is really new in this contribution is unclear while reading.

Ans: We thank the reviewer for raising this important point. The novelty of this study lies in providing a full-water-column assessment of POM sources on a continental shelf, with particular emphasis on quantifying the terrigenous contribution. In the revised manuscript, we clarify the motivation for this work by highlighting the difficulty in explaining the occurrence of ¹³C-depleted POM frequently observed in subsurface and bottom shelf waters. We have added a paragraph in Introduction to explicitly describe this problem.

[2] Moreover, why the authors have used 2 step estimation of terrigenous OC using lignin and δ¹³C of organic carbon and how it is understood that lignin is appropriate than δ¹³C and in what capacity.

Ans: We thank the reviewer for this insightful comment. We adopted the two-step approach for the following reasons.

1. Similar to nearby continental shelves, the northeastern Taiwan Strait contains POM with relatively low δ¹³C values in subsurface and bottom waters. Identifying sources of this POM is an important component of our goal to provide a full-water-column assessment of OM sources.
2. The order in which the proxies are used in the two-step approach does not imply that one proxy is superior to the other. Lignin was used in Step 1 because the ambiguity primarily arises from δ¹³C signatures. Lignin phenols are diagnostic biomarkers of vascular plants and therefore provide an independent constraint on the presence of terrigenous OM. The underlying hypothesis is that if the low-δ¹³C POM has a terrigenous origin, it should contain elevated Λ8 (OC normalized lignin concentration) relative to other samples. Such relationships between lignin abundance and δ¹³C_OC have been widely documented in sedimentary OM (Bianchi et al., 2018).
3. Although the lignin data were useful for diagnosing the relative abundance of terrigenous POM (we found that most low-δ¹³C POM samples contained negligible lignin), it remained challenging to quantify terrigenous POC (POC_terr) directly from lignin. The main difficulty lies in the uncertainty in the ratio of lignin to total terrigenous OM in the source material (Λ8_terr,s). This motivated Step 2, in which we quantified POC_terr in high-TSM samples using a δ¹³C-based mixing approach. These high-TSM samples were strongly affected by resuspended seabed sediment based on their spatial distribution, which allows more reliable assignment of endmember δ¹³C values.
4. The results from Step 2 provide not only an independent estimate of POC_terr, but also Λ8_cal (the source signature of Λ8 required to account for the measured Λ8 in POM samples) that can be compared with the Λ8_terr,s used in Step 1. This comparison helps evaluate the limitations and uncertainties of lignin-based POC_terr quantification.

To clarify this rationale, we have revised Sect. 3.3.1 (Estimation of POC_terr​) in the manuscript.

[3] In their previous publication, Lin et al. (2025a), the authors described source to sink processes, OM composition and oxygen consumption. This should be categorically explained to understand the importance of already published one from the present manuscript content, not to mention that the hydrographic, POM and carbonate chemistry data for the present study were taken from Lin et al. (2025b). My question is why not the authors can focus on new, unpublished data to revise the manuscript?

Ans: We added a sub-section "3.1 Data sources and previously published data" to clarify which part of the dataset has been used in Lin et al. (2025a), and which part of the dataset is new.

[4] After using lignin and δ¹³C, the authors mentioned in lines 300-305 that “the available evidence does not allow us to resolve which process is chiefly responsible for the low Λ8 values”. This part shows some incorrect handling of data and interpretation. The authors should know what their data can or cannot reveal; after so many corrections and recalculations and re-estimations in the manuscript, such a vague statement bespeaks that the authors have no clear-cut idea to understand or reconcile lignin and δ¹³C data in this manuscript. This needs more vigorous approach of data reconciliation and that is missing in the manuscript.

Ans: We thank the reviewer for raising this important point. We checked the literature again and realized that we misinterpreted the data of Wakeham et al. (2009). In this paper, the notation "Λ_m" represents lignin concentration per unit dry weight of sediment, whereas in most literature (e.g., Bianchi et al., 2018; Hedges et al., 1997; Hernes and Benner, 2006) and our manuscript, and Greek letter Λ denotes OC normalized lignin concentration. We calculated the OC normalized lignin concentration (Λ8 in the table below) of density fractions presented in Wakeham et al. (2009, https://doi.org/10.1016/j.marchem.2009.08.005) (please see the attached file for the table).

We found that the low-density fractions, while enriched in lignin on a dry weight (dw) basis, exhibited lower Λ8 relative to the bulk sediment (bold text in the table). This is because OC was enriched more strongly in low-density fractions than lignin. Our result of "Λ8_cal averaged only 17 ± 10 % of Λ8_terr,s", where Λ8_terr,s represents the value of riverine or seabed sediment and Λ8_cal the theoretical value of source material that can explain the measured value in POM, is consistent with their observation. However, because the reduction in Λ8 of low-density fractions relative to bulk sediments was highly variable and appeared site-specific, we did not attempt a detailed correction of the lignin-based POC_terr, but simply provided a first-order assessment of the potential magnitude of the bias. We have revised the relevant paragraph in a new section (Sect. 5.1.2, Uncertainty in POC_terr estimation).

MINOR COMMENTS

[5] Section 2.1, Lines 75-80: The authors mentioned that the shelf off SE China or through the funnel-shaped Penghu Channel, which serves as the primary pathway for volume transport (Jan et al., 2002). If any estimate on the volume transport is available, it is better to include here with reference(s).

Ans: We have added the number of volume transport through the Penghu Channel in summer to the revised manuscript.

[6] In the same paragraph, it is mentioned that tidal current velocities decreased progressively. Any estimate of tidal velocities may be included here.

Ans: We have added the relevant numbers to the revised manuscript.

[7] Line 143: Change to Andrew

Ans: We have corrected the spelling.

[8] Lines 295-310: Terrigenous POM persists during both alongshore and cross-shelf transport. Any reason from biomarker data why land-derived POM survives during the both transport?

Ans: We have clarified this point in the revised manuscript (Sect. 5.1.1) using lignin biomarkers. Higher (Ad/Al)_V ratios offshore indicate longer exposure to degradation in SCSSW than in nearshore TCW, consistent with longer transit times (~15 vs. ~5 days). However, both timescales are short relative to the slow degradation kinetics of lignin (Benner et al., 1987). This mismatch explains why lignin, and thus a fraction of terrigenous POM, can persist during both alongshore and cross-shelf transport.

[9] The authors said that “Seabed sediments on this region are dominated by terrigenous OM (Lin et al., 2025a)”. However, they also mentioned that “subsurface shelf waters contained negligible lignin consistent with SCSW receiving a greater contribution from Pacific-origin waters. From this line, I understood that negligible lignin is derived from Pacific waters. The study area is proximal to both Taiwan and mainland China terrestrial fluxes that dominated likely by lignin. Given this, the above statement seems to me “odd” in the study area, which is shallow and dynamic strait.

Ans: We have clarified this point in the revised manuscript (Sect. 5.1.1). Although the study area receives substantial terrigenous input, the low lignin concentrations in offshore subsurface waters can be explained by source water mass and supply limitations. These waters are primarily derived from SCSW, which is likely lignin-poor due to a stronger contribution from Pacific-origin waters. In addition, weak resuspension under calm sea-state conditions limited the upward transport of sediment-derived lignin, while low river discharge confined plume influence to a narrow coastal zone, reducing lignin supply to subsurface layers. These factors together explain why lignin can be minimal in subsurface waters despite strong terrigenous influence in sediments.

[10] In Fig. C6, why just one data has not been excluded in the regression analysis, though given r² value is low?

Ans: We thank the reviewer for raising this important point. In the original analysis, one data point was excluded from the regression based on its apparent deviation from the overall data distribution. Recognizing that such subjective exclusion may be contentious, we have retained this data point in the revised analysis. We have updated the statistical results in Fig. C6 and Sect. 5.2.2 (POC/Chl ratios track photoacclimation). With the full dataset included, the correlation between POC and Chl for the DCM samples is no longer significant. Therefore, instead of relying on regression-derived overall POC/Chl ratios, we now compare nearshore waters and DCM layers using their measured POC/Chl ratios. The difference between the two groups is evaluated using a Mann-Whitney test, which shows that nearshore waters exhibit significantly higher values than DCM layers.

This is to submit the attachment missing in the previous reply. 

We appreciate the reviewer’s thoughtful evaluation of our manuscript and the valuable suggestions provided. These comments have helped us improve the presentation of our results. We have carefully considered each point and revised the manuscript accordingly, with detailed responses provided below. A rebuttal with screenshots of the revised text in the manuscript is included in the attached file.

[1] Authors have already published sedimentary POM assessments in the same region in Lin et al. (2025). I recommend clarifying new findings in this paper, which is different from the previous study. Although several datasets were derived from the previous study, it is sometimes unclear which data is newly measured.

Ans: We added a sub-section "3.1 Data sources and previously published data" to clarify which part of the dataset has been used in Lin et al. (2025a), and which part of the dataset is new.

[2] Authors applied two step estimation of terrigenous POC based on lignin (Σ8) data and δ¹³C_OC mixing model. Σ8 was primarily sourced by resuspended sediment because of stronger correlation with TSM than salinity. However, Σ8 was also transported by riverine input. Thus, are there lignin preserved in sediments and fresh lignin transported by rivers? Did the differences in lignin affect the subsequent calculations of terrigenous POC?

Ans: We thank the reviewer for this insightful comment. In the original manuscript, we adopted Λ8_terr,s values from either riverine TSM or seabed sediments. In the revised manuscript, we now acknowledge that there may be a mixed contribution from these two sources. Such mixing would lead to overestimation of POC_terr in surface waters and underestimation in subsurface and bottom waters. Given the larger volume of subsurface and bottom waters, our overall estimate is therefore likely biased low. We did not apply a detailed correction due to the lack of quantitative constraints on the relative contributions of these sources.

[3] After L301, authors assessed lignin-based estimates (underestimation) and ¹³C_OC-based mixing model (overestimation). Which estimation is better? So, is it the conclusion that “terrigenous POM has a limited contribution in shelf water, but it is main component in seabed sediments” for this paragraph?

Ans: We thank the reviewer for this question. In the revised manuscript (Sect. 5.1.2), we clarify that the δ¹³C-based mixing model provides more reliable estimates for high-TSM samples, as these are primarily derived from seabed sediments that contain a high terrigenous OM content.

For low-TSM samples, however, δ¹³C_OC is less suitable as a tracer due to ambiguity in interpreting low-δ¹³C_OC signatures. The lignin-based estimates are likely biased low overall, and a first-order assessment using the average Λ8_cal/Λ8_terr,s ratio suggests that f_terr could increase by up to sixfold. Under this scenario, terrigenous OM would still represent a secondary component in these samples. Therefore, our results support the conclusion that terrigenous POM makes only a limited contribution to shelf waters above the benthic nepheloid layer.

The conclusion that the terrigenous source is the main component of seabed sedimentary OM is based on Lin et a. (2025a), not the present study.

[4] Authors suggest “terrigenous correction can be reasonably omitted in future analyses of low-TSM shelf waters, even without lignin data” in L330-331. If the correction did not alter the POM features, POM source is estimated primarily to be marine organisms. Please clarify the importance of accurately estimating the terrigenous POC contribution which is quite small.

Ans: We thank the reviewer for this insightful comment. We agree that terrigenous POM represents only a minor fraction of bulk POM in low-TSM shelf waters, and that its removal does not significantly alter the overall biogeochemical characteristics of POM. This is now clarified in Sect. 5.2, where we show that terrigenous POM exerts limited influence on bulk properties and that the main conclusions are insensitive to whether the correction is applied.

However, accurately constraining POC_terr remains important for understanding its transport and fate. As revised in Sect. 5.1.3, even though the fractional contribution of terrigenous OM is small, low-TSM shelf waters can account for a substantial portion of its lateral export, particularly if the potential underestimation associated with the lignin-based approach is taken into account. This indicates that low-TSM waters, despite their low terrigenous signal, can play a non-negligible role in advective transport.

We have revised the manuscript to better distinguish between these two aspects, namely the limited role of terrigenous POM in controlling bulk POM properties and its potentially important role in regional carbon transport.