the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Oct 2026
Tracing Sediment and Phosphorus Source Hysteresis During an Extreme Flood in a Subtropical Australian River
Abstract. Extreme flood events are increasingly recognised as important drivers of suspended sediment (SS) and particulate phosphorous (PP) transport. However, their dynamics remain poorly understood due to the stochastic distribution and limited high-frequency sampling of these events. This study is the first to combine high-resolution rising-stage sampling, hysteresis analysis, and a novel sediment fingerprinting approach based on elemental fractions to quantify the changes in SS and PP sources during the 2025 Cyclone Alfred flood in a subtropical Australian catchment. Results show a pronounced figure-eight hysteresis in both the SS and PP concentration-discharge relationship, characterised by an early first flush and a secondary peak at highest discharge. Sediment tracing shows a shift in source contribution, with initial dominance of alluvial bank material (60–80 %) transitioning to hillslope sources (~70 %) near peak discharge. Despite its short dominance, hillslope material contributed ~60 % of the total storm event sediment load, compared to ~30 % from alluvium sources. In parallel, PP shifted from predominantly oxide-bound forms during the rising limb (~44–75 %) to exchangeable (bioavailable) at peak flow (~70 %), wherein the latter contributed ~53 % of total event PP export. The Cyclone Alfred event alone accounted for ~57 % of annual sediment yield and ~65 % of annual particulate P export. Extreme events thus non-linearly activate distinct SS and reactive PP sources and dominate yearly PP and SS loads. These findings highlight the need to extend monitoring and management strategies to hillslope sediment mobilisation and identifying the sources of nutrient pulses under intensifying urban development and climate extremes.
- Preprint
(1384 KB) - Metadata XML
-
Supplement
(2173 KB) - BibTeX
- EndNote
Status: open (until 17 Nov 2026)
- RC1: 'Comment on egusphere-2026-4743', Anonymous Referee #1, 13 Oct 2026 reply
Data sets
Data for Manuscript: Tracing Sediment and Phosphorus Source Hysteresis During an Extreme Flood in a Subtropical Australian River M. Wynants et al. https://doi.org/10.5281/zenodo.21816185
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 236 | 69 | 27 | 332 | 37 | 32 | 29 |
- HTML: 236
- PDF: 69
- XML: 27
- Total: 332
- Supplement: 37
- BibTeX: 32
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Abstract
The abstract clearly communicates the novelty, methods and principal findings.
The statement that hillslope material contributed approximately 60% of the event sediment load should be accompanied by an indication of uncertainty. The result depends on the selected tracer cluster, mixing-model assumptions and interpolation across the unsampled peak.
The estimates that the event represented 57% of annual sediment yield and 65% of annual particulate P export are presented too definitively. These are scenario-based extrapolations derived by applying relationships from a single event to the annual hydrograph and should be described as provisional estimates.
Introduction
The Introduction provides a strong regional and methodological rationale, but it could state the core knowledge gap more concisely. The main gap appears to be the lack of direct evidence linking event-scale hysteresis patterns to temporally changing sediment sources.
The limitations of existing fallout-radionuclide and geochemical approaches are somewhat broadly stated. Fallout radionuclides may differentiate surface and subsurface sources effectively even if they cannot distinguish all geomorphic source types. The wording should avoid implying that these approaches are generally inadequate.
The hypothesis is currently broad: “extreme events non-linearly activate different SS and PP sources.” It should be converted into explicit, testable predictions, for example: alluvial material will dominate the early first flush; hillslope contributions will increase above a discharge or rainfall-connectivity threshold; phosphorus fractionation will change alongside sediment-source contributions.
Materials and methods
Reconsider the phosphorus terminology. Evaporating the entire water sample before extraction means that dissolved phosphorus was retained and subsequently measured, especially in the first extraction. Results from this procedure should not be labelled solely as particulate P. The paper should either revise the operational definitions throughout or provide evidence that dissolved P was negligible.
Address temporal and spatial representativeness of source samples. Source materials were collected approximately six to seven months after the flood, and some source categories had small and uneven sample numbers, including five hillslope subsoil and three upland subsurface samples. Explain whether post-event erosion, deposition or land disturbance could have altered source signatures, and assess how unequal sample sizes affect model uncertainty.
Verify the hysteresis-index equation. The authors should check the equation, code and reported HI values. The handling of bins containing only rising- or falling-limb observations should also be explained.
Strengthen uncertainty and validation procedures. Report analytical precision, certified-reference-material recoveries, extraction blanks, replicate variability and the number of values below detection limits.
Results
Results should be reported with uncertainty wherever possible. Source contributions such as “around 70%” and load estimates such as 90,299 tonnes and 199 tonnes appear overly precise given mixing-model uncertainty, concentration interpolation and the missing peak sample.
Statistical results should report sample size, correlation coefficient and exact p-value consistently. Statements such as“p-value ~0” should be replaced with the calculated value or a threshold such as p<0.001p<0.001p<0.001. Multiple testing should also be acknowledged or corrected for.
Discussion
The proposed sequence of alluvial first flush, temporary supply limitation and later hillslope connectivity is plausible and supported by converging evidence. Nevertheless, phrases such as “confirmed” should be replaced with wording such as “consistent with”, because the study lacks direct observations of erosion and connectivity across the catchment.
The attribution of the peak exchangeable-P pulse to wastewater, slurry-storage overflow or other point sources is speculative. No point-source samples, dissolved-P measurements, microbial indicators or wastewater tracers were collected. These should be framed as hypotheses for future testing, not likely explanations.
The paper equates the operationally extracted exchangeable fraction with bioavailable P. Bioavailability depends on receiving-water chemistry, desorption kinetics, biological uptake and transformation during transport. The authors should use “potentially labile” or “operationally exchangeable” unless direct bioavailability measurements are available.
The argument that urban development near Canungra contributed to hillslope sediment export is important but unsupported by the three-source mixing model, which combines multiple hillslope land uses. Avoid attributing the hillslope signal specifically to urban development unless further land-use-level discrimination or independent evidence is provided.
The annual extrapolation should be repositioned as an exploratory scenario rather than a sediment budget. Applying concentration and source relationships from one extreme event to all hydrological states assumes temporal stationarity and may not represent antecedent conditions, seasonal sediment availability or smaller storms. A sensitivity envelope and stronger qualification are needed.
Conclusions
The Conclusion provides a clear synthesis, but the opening statement that the hypothesis was “confirmed” is too strong. “Supported by the observations and model results” would better reflect the limitations.
The conclusion that hillslope material contributed most event-scale sediment export should include uncertainty and acknowledge the missing peak sample and selected mixing-model solution.
The text again refers to exchangeable P as inherently bioavailable. Revise this to an operationally defined labile fraction that may include dissolved P.
Potential urban and agricultural point sources should be identified as untested explanations requiring targeted monitoring rather than inferred event sources.
The final paragraph could be shortened and focused on three defensible contributions:
high-frequency sampling revealed complex event hysteresis; source apportionment indicated a transition from alluvial to hillslope sediment; and extreme-event monitoring is necessary to understand annual sediment and phosphorus export.
Figures
Figure 1: The catchment map contains useful information. I would revised the extent map to make it more obvious where the sample site is in Aus. The text alternates between “Bromfleet” and “Blomfleet”; this should be corrected.
Figure 4: Caption- The unit should be written as µm, not “um”.
Figure 5: The annual-budget figure may visually imply greater certainty than the extrapolation permits. Label it explicitly as a scenario-based annual estimate, add uncertainty ranges, and separate the observed Cyclone Alfred contribution from modelled contributions associated with other events and baseflow.