Sample preparation was guided by Glaser et al. (1998) and Dittmar (2008). To prepare samples for BPCA analysis, each extracted DOM sample (2 mL) was inserted into glass ampules. DOM samples were dried completely under a stream of N₂ gas at room temperature. Once dried, each ampule was rinsed with 5 mL HNO₃ (70%) and put directly into thoroughly sanitised Teflon reaction vessels (CEM Mars Xpress). Samples were digested for 6 h at 170 °C. Following oxidation, samples were allowed to cool completely and then pipetted into borosilicate vials. Each sample was then evaporated under a N₂ stream at 50 °C. Samples were wrapped in foil and stored in the freezer until BPCA analysis.

BPCA analysis was conducted using a modified method based on Hindersmann and Achten (2017). The BPCA analysis was performed using an Agilent 1,290 Infinity II Liquid Chromatography system coupled with a Quadrupole Time-of-Flight mass detector (LC-QTOF -MS). For liquid chromatography, a Phenomenex Kinetex EVO C18 Column (1.7 μm, 100 Å, 100 × 2.1 mm) was used as the stationary phase and the column temperature was 20 °C. The mobile phase contained two parts, A and B, and the total flow rate of the mobile phase was 0.3 mL/min. The mobile phase A was 100% milli-Q water containing 4% formic acid, and the mobile phase B was 100% acetonitrile. A gradient elution was used: 0–5 min 100% A (0%B), 5–8 min 0–8% B, 8–11 min, 8–12% B, 11–11.3 min 12–99% B, 11.3–11.6 min 99% B, 11.6–12 min 99-0%B, 12–16 min 100%A (0%B). Samples were injected in duplicate, and the injection volume was 1.0 μL.

For mass spectrometry, the ionisation was carried out via an Agilent Dual Jet Stream electrospray source. The mass spectrometry was operated in negative mode with the drying gas flow at 10 L min⁻¹ N₂, at 250 °C; sheath gas flow at 12 L min⁻¹ at 250 °C; nebuliser pressure at 20 psi; Vcap voltage at 3,000 V, Fragmentor at 50 V; Skimmer at 45 V, Qct 1 RF Vpp at 750 V. The TOF mass range was between 50 and 1,700 m/z, and the acquisition rate was 2 spectra/s.

1,2,3-Benzenetricarboxylic acid (B3CA, Product No. B4201), 1,2,4,5-benzenetetracarboxylic acid (B4CA, Product No. B4007), mellitic acid (B6CA, Product No. M2705) were used as the standards to identify and quantify BPCA. Three B3CA isomers, three B4CA isomers, B5CA and B6CA were identified by comparing the retention times of pure standards and accurate mass-to-charge ratios (m/z). B3CA was quantified based on deprotonated molecular ion [M–H]⁻, 209.0102 m/z at 5.283 min, 6.960 min, and 9.680 min. B4CA was quantified based on deprotonated ion [M–H]⁻, 253.0018 m/z at 1.800 min, 2.100 min, and 3.878 min. B5CA was quantified based on deprotonated and dehydrated deprotonated [M–H–H₂O]⁻ molecular ion, 234.9886 m/z at 1.110 min. B6CA was quantified based on deprotonated and dehydrated deprotonated [M–H–H₂O]⁻ molecular ion, 322.9693 m/z at 0.790 min. Quantification of BPCA was carried out by external calibration standards based on the area under peaks of extracted ion chromatograms for BPCAs (ions listed above). Linear calibration curves were established between 0.5 and 80 μg mL⁻¹ for pure BPCA. The available B3CA and B4CA standards were used to quantify their isomers. As no commercial standard was found for B5CA, B5CA was quantified based on the calibration curve of the available B4CA standard.

Dissolved black carbon was calculated based on the concentrations of the individual BPCAs (mol BPCA L⁻¹), as provided in Dittmar (2008): DBC (mol C L⁻¹) = 33.4 × (B6CA + B5CA + 0.5 B4CA + 0.5 B3CA).

Surface water DOC concentrations ranged from 145 to 540 μmol L⁻¹, with the lowest values occurring at Burrill (MEAN: 183.23 ± 20.75 μmol L⁻¹) and the highest at Termeil (MEAN: 540.42 ± 0.42 μmol L⁻¹) (Figure 2; Appendix Table 3). PERMANOVA confirmed a significant difference in the surface water DOC concentrations of the different estuaries (Pseudo-F = 43.293, P(perm) = 0.0001, unique 9,960 permutations). Pairwise tests revealed DOC in surface water was significantly higher in Termeil than Wonboyn, which was significantly higher than all other estuaries except Durras, which has significantly higher DOC than Conjola, which is significantly higher than Tuross and Burrill (Table 2).

Shallow groundwater DOC was markedly higher than surface water across all sites (Table 2), ranging from 1,308.3 to 3,043.3 μmol L⁻¹. The highest concentration was observed at Wonboyn (3,043 μmol L⁻¹), and the lowest at Termeil (1,308 μmol L⁻¹). Marginal tests in the DBLM indicated that the water source (surface water vs. shallow groundwater) was the only significant predictor independently, explaining 66.3% of the variation in DOC on its own (p = 0.001) (Appendix Table 4).

The BEST model, which explained 86.94% of variation in DOC concentrations across all samples (AICc = 362.8), included catchment area (km²), percent of high severity burn, and water source (surface water vs. groundwater) as predictor variables. A closely performing alternative model explained 86.6% of the variation and included moderate severity burn, catchment area, and water source. Low-severity burn was excluded from the DBLM due to high collinearity with other burn categories, specifically it was negatively correlated with extreme burn (r = −0.942) and positively correlated with moderate burn (r = 0.929) (Appendix Table 2).

Surface water DOC concentrations observed in this study align with reported ranges in post-fire aquatic environments (Dittmar et al., 2012a,b; Wei et al., 2021). For example, Dittmar et al. (2012a) reported surface water DOC concentrations ranging from 229 to 608 μmol L⁻¹ in the lower creeks and river tributaries of the Apalachicola Bay intertidal marsh system in Florida.

While this range overlaps with the DOC concentrations observed in this study (145–540 μmol L⁻¹), it is important to note that their samples were collected 3 years after the fire event, suggesting that DOC concentrations in the present study, taken within a shorter post-fire timeframe, may be relatively low by comparison. Furthermore, some of the highest concentrations in Dittmar et al. (2012a) were recorded in tidal creeks, where restricted flow and limited tidal exchange likely contributed to DOC accumulation. In contrast, the estuaries in this study, particularly Tuross, likely experienced greater flushing, leading to more diluted DOC levels. Additionally, prescribed burning is a common land management practise in the Apalachicola region, potentially influencing DOC inputs differently than in the wildfire-affected systems explained here.

DOC concentrations have typically been found to decline from headwaters to coastal waters due to dilution, photodegradation and flushing (Jones et al., 2016; Moody et al., 2013; Crosswell et al., 2017; Moyer et al., 2015; Dittmar et al., 2012a). While our results demonstrated relatively uniform concentrations throughout estuary surface water—showing no clear upper to lower estuary gradient at most sites—these processes are still likely to influence DOC levels at the broader estuary scale. Specifically, lower DOC concentrations were generally observed in larger, more flushed estuaries (e.g., Tuross, Burrill, and Conjola), while higher concentrations occurred in smaller systems with reduced flushing and lower water volumes (e.g., Wonboyn, Termeil, and Durras). This suggests that although longitudinal variation of DOC within estuaries may be limited, dilution, photodegradation, and flushing remain important drivers of DOC variability across systems.

Differences in DOC concentrations between estuaries also likely reflect of a combination of catchment features, such as the catchment size, proportion of vegetated area, fire severity and recovery stage. For example, Termeil—a small ICOLL with limited ocean exchange, a predominantly forested catchment (NSW Government, Environment and Heritage, 2021), low water levels at the time of sampling, and mostly low to moderate severity burn (45.40, 38.84%, respectively) (NSW DCCEEW, 2024) (Table 1)—had the highest surface water DOC concentrations. In contrast, Tuross experienced extensive high and extreme severity fire (37.5, 42.3%) within its catchment, which has been observed to lead to decreased carbon stocks in organic soil horizons in comparison to unburned soils (Maestrini et al., 2017; Adkins et al., 2019). While intense fire can initially mobilise large pulses of organic carbon, long-term effects often include sustained reductions in DOC availability. Santos et al. (2019) identified a significant interaction between fire severity and time since fire in shaping DOC concentrations in fire-affected streams, showing that high intensity burns can lead to prolonged decreases in DOC lasting well into the recovery period—a duration that varies depending on the specific ecosystem (Wei et al., 2021). In contrast, the study also found that low severity fires, where canopy remains intact, can result in increased DOC concentrations in streams over time. Given the extensive high severity burn that occurred in Tuross’s catchment, it is likely that large pulses of organic matter were mobilised during and after the immediate post-fire rainfall events (Johnston and Maher, 2022). This would have been followed by significant reductions in soil DOC, which combined with increased flushing, would have ultimately led to DOC concentrations in surface waters (Santos et al., 2019). However, without pre-fire baseline data, it is difficult to be sure whether these levels represent an increase/decrease, recovery or continuation of pre-existing conditions.

Surface water DBC concentrations ranged from 2.33 to 38.86 μmol L⁻¹ (Figure 2). The highest averages were measured at Wonboyn (23.11 ± 4.29 μmol L⁻¹), followed by Termeil (21.17 ± 3.74 μmol L⁻¹), Durras (13.36 ± 4.48 μmol L⁻¹), Conjola Lake (6.47 ± 1.58 μmol L⁻¹), Tuross (6.27 ± 0.79 μmol L⁻¹), and lowest at Burrill (3.79 ± 0.58 μmol L⁻¹) (Figure 3). PERMANOVA confirmed a significant difference in the surface water DBC concentrations of the different estuaries (Pseudo-F = 9.3042, P(perm) = 0.0002, unique 9,953 permutations) (Table 2). Marginal tests in the DBLM indicated that source of sample water (shallow groundwater vs. surface water) was the only predictor of DBC concentrations that was significant on its own, explaining 66.3% of variation in concentrations (p = 0.0001). The BEST model included high burn (%), moderate burn (%), and source of the water sample to explain 76.5% of variation in DBC concentrations (AICc = 164.68). The addition of extreme burn (%) and estuary volume to the above model explained 79.5% of variation in DBC (AICc = 166.07) (Appendix Table 4).

Wonboyn estuary recorded the highest DBC concentrations, consistent with elevated DOC levels, ranging from 14.27 μmol L⁻¹ (3.66% of DOC) to 38.86 μmol L⁻¹ (8.43% of DOC) (Appendix Table 4). Pair-wise tests showed significant differences between Wonboyn surface water DBC concentrations and those found at Tuross, Conjola, and Burrill (Table 2). Wonboyn experienced the most extensive catchment burn out of the studied estuaries (Table 1), with approximately 91% of its 335 km² catchment burned, much of which was classified as “high severity” burn (Table 1) (DPE, 2020).

In most estuaries, lower surface water DBC concentrations were observed near tidal exchange zones, reflecting dilution, reduced residence times, and the influence of riverine inputs. However, this gradient was not evident in Wonboyn, where surface waters near the estuary mouth retained relatively high DBC. This likely reflects the unique influence of adjacent fire-affected areas. During the 2019/2020 fires, the burn extended to the estuary’s edge, burning through much of the saltmarsh habitat, likely providing a direct and ongoing source of PyC to estuarine waters via lateral transport (NSW DCEEW, 2020; Dittmar et al., 2012a). Elevated DBC concentrations in the upper estuary near freshwater inflows also point to substantial inputs from upstream tributaries.

By contrast, Burrill exhibited significantly lower surface water DBC concentrations, as confirmed by PERMANOVA and aligning with its lower DOC levels. Burrill’s catchment was the least affected by the Black Summer fires (20% burned, <8.5% at high or extreme severity), which likely influenced its relatively low DBC values (Table 1; Figure 3). Tuross also showed comparatively low surface water DBC concentrations, despite experiencing extensive fire across its large catchment (87% of its 1,574 km² burned, including 44.33% at extreme and 37.46% at high severity) (Table 1). In this case, other factors appear to be driving DBC patterns. Tuross has a permanently open mouth, promoting increased tidal flushing and shorter residence times, which likely reduced DBC accumulation. In addition, because most of the burned area was located upstream rather than adjacent to the estuary, overland inputs were buffered, and DBC transported via river flow likely travelled greater distances before entering the estuary (Barros et al., 2022). During transit, significant losses may have occurred through photodegradation (Stubbins et al., 2010), deposition, or transformation into particulate forms that settled out before reaching the estuary (Xu et al., 2017). Also, chars produced at high temperatures can result in highly condensed forms of PyC, which are less soluble and therefore less likely to make it into downstream aquatic systems (Schneider et al., 2010; Abiven et al., 2011; Barton et al., 2024). These processes may have significantly reduced DBC concentrations in the estuary, despite the catchment’s extensive high severity burn.

The estuary-specific differences observed illustrate the complex and non-linear relationship between fire severity, catchment burn extent, and surface water DBC concentrations. While intense fires tend to produce more pyrogenic material (Komarek et al., 1973; Czimczik et al., 2005), downstream concentrations depend heavily on catchment characteristics, transport dynamics, and estuarine processing. For example, Tuross, despite being one of the most severely impacted catchments, displayed relatively low surface water concentrations of both DOC and DBC. This pattern may reflect post-fire reductions in carbon stocks (Maestrini et al., 2017; Adkins et al., 2019), as well as enhanced flushing and carbon losses during transport to the estuary. While fire severity is likely to be a key driver for initial DOC and DBC levels in surface waters, their export dynamics may differ depending on catchment recovery, connectivity to estuarine waters, and estuarine processing mechanism.

Shallow groundwater DBC concentrations ranged from 22.93 to 77.80 μmol L⁻¹ (Appendix Table 3). The highest levels were observed at Wonboyn, consistent with surface water results, and the lowest were found at Termeil. Shallow groundwater DBC was generally higher than surface water concentrations across all estuaries, with the exception of Termeil, at which a single surface water site was slightly higher than shallow groundwater levels. The elevated shallow groundwater found at Wonboyn was likely due, in part, to the extensive burn occurring through estuarine and tidal marsh areas. The elevated DBC levels found in one of the shallow groundwater sites, located in the tidal marsh, as well as in the adjacent surface water site (Figure 3), indicate that groundwater discharge from the marsh contributes to DBC levels in surface waters. This is consistent with results from prior studies demonstrating tidal flushing as an important pathway for organic matter transfer from intertidal areas to estuaries (Dittmar et al., 2001, 2006, 2012a). Dittmar et al. (2006) estimated that >10% of the world-wide land-ocean DOM flux originates from mangroves, suggesting that material transport from marshes could be similarly efficient (Dittmar et al., 2006; Dittmar et al., 2012a). Although further study is required, we propose that tidal pumping and shallow groundwater discharge from the marsh appear to be important sources of DBC in the estuary.

Tuross shallow groundwater was among one of the higher DBC values observed, despite being sampled from a small, vegetated island towards the centre of the estuary, which was not directly impacted during the 2019/2020 fires (Figure 3). Due to the sample site’s location (Appendix Table 1) and distance from the fire-impacted catchment, DBC likely accumulated from tidal flushing and deposition of pyrogenic material to the island from the upstream burnt catchment. Such material could potentially leach DBC into shallow groundwater, acting as a source of DBC to estuarine waters over time.