Distribution pattern of dissolved organic matter in pore water of sediments from three typical areas of western Lake Taihu and its environmental implications Open Access

As a relatively reactive component of the organic matter (OM) pool in aquatic ecosystems, dissolved organic matter (DOM) comprises thousands of degradation by-products as well as freshly produced compounds with a molecular weight (MW) ranging from <1,000 to 100,000 Da (Kellerman et al. 2015; Li et al. 2022). The migration, transformation, and accumulation of DOM in pore water of sediments are pivotal links of biogeochemical processes related to lacustrine carbon cycling (Ziegelgruber et al. 2013; Mostofa et al. 2018; Li et al. 2021). The horizontal-spatial characteristics of quality and composition for DOM in pore water can reveal the relative contribution levels from terrestrial (allochthonous) and aquatic (autochthonous) sources of organic carbon and their input mechanisms in lake ecosystems (Chen & Hur 2015; Derrien et al. 2017; Zhu et al. 2022). Vertical changes of DOM in pore water of sediment cores are critical indicators for the fate of massive organic carbon deposited in sediments, the biogeochemical effects associated with the efflux and influx of dissolved organic carbon (DOC) and dissolved inorganic carbon (DIC) in bottom water, and the evolution information of lake and their basins (Yang et al. 2014; Li et al. 2021; Wu et al. 2022). Therefore, clarification on the distribution pattern of DOM in pore water and its environmental implications in lake ecosystems is essential to understand the driving force of human activities and natural factors on the lake carbon cycle.

In eutrophic lakes, a large amount of DOM from multiple sources accumulates in the sediment core, including terrestrial DOM that enters through inflow rivers and autochthonous DOM released by phytoplankton (especially, algae), macrophytes, and microbial processes (Li et al. 2021; Wen et al. 2022). The degradation, migration, and transformation of DOM from various sources present obvious differences in water columns and sediments due to their diverse physicochemical properties as well as composition and molecular size in lake ecosystems (Zhou et al. 2021; Lee et al. 2023). For example, autochthonous DOM containing abundant protein and labile polysaccharides is readily biodegradable while terrestrial DOM consisting of more humic substances and structural polysaccharides (e.g., cellulose) is recalcitrantly depredated by bacteria (Perez & Sommaruga 2006; Guillemette et al. 2016). Meanwhile, human disturbance (e.g., dredging works) and natural environmental conditions (e.g., water temperature, wind waves, oxidation–reduction state) affect the gradient distribution of accumulated DOM with depths, in pore water of sediments, causing changes in the diffusion direction and flux size of DOC across the water-sediment interface (Yang et al. 2014; Derrien et al. 2017; Li et al. 2018). Especially, the dredging work of sediments in eutrophic lakes can cause the removal of fresh sediments or the mixing of fresh and older sediments, which probably changes the gradient characteristics of DOM due to the disturbance of dissolved oxygen (DO), pH, nutrient levels, and microbial community structure in post-dredging sediments (Zhang et al. 2010; Wan et al. 2020). Currently, the study of DOM in eutrophic lakes generally centers on the sources, structures, and fates in overlying water and pore water of surface sediments as well as its biogeochemical links with the trophic state, succession of aquatic organisms, and pollutant behaviors (Jiang et al. 2018; Zhang et al. 2021; Wen et al. 2022; Bao et al. 2023). However, few comparative studies have been performed on the distribution of DOM in pore water of sediment cores from the typical lake areas of different ecological types excessively disturbed by human activities.

At present, ultraviolet–visible (UV–Vis) absorption spectrometry and fluorescence emission–excitation matrix spectrometry (EEMs) are the most widely applied for exploring DOM compositions, sources, distribution, and molecular sizes due to great convenience, highly rapid sensitivity, and low cost. Based on absorption spectroscopy, absorbance, absorption coefficient, and absorbance ratio are used to characterize the concentration or abundance and chemical composition of DOM (Helms et al. 2008; Li & Hur 2017). EEM is applied for distinguishing the sources, components, and composition of DOM based on different classes of fluorescence compounds and the four main indexes of EEM coupling with PARAFAC (i.e., FI, FrI, BIX, and HIX) (Osburn et al. 2012; Zhang et al. 2021). Moreover, the benthic fluxes of DOC can be estimated by Fick's first law based on the concentration gradients of pore water DOC in lacustrine and marine sediments (Burdige et al. 1992; Chen et al. 2017; Li et al. 2018). Thus, these methods and techniques in aquatic ecosystems are powerful tools in reflecting the depth-dependent properties and vertical dynamics of DOM in pore water of sediments, and further estimating organic carbon fluxes in large shallow eutrophic lakes.

Lake Taihu is the third largest shallow freshwater lake in China, covering an area of 2,338.1 km² with an average depth of 1.9 m (Lu et al. 2019). In recent decades, eutrophication has been one of the most serious environmental problems in Lake Taihu, exhibiting excessive algal blooms and overall deterioration of water quality (Liu et al. 2017; Lai et al. 2023). In order to control the eutrophication of Taihu Lake, the government and researchers have explored various in situ ecological restoration methods and engineering techniques, such as constructing macrophytes belts and dredging engineering (Chen et al. 2021; Yin et al. 2021). When the carbon cycling of Lake Taihu is influenced by the eutrophication effect, inputs of inflow rivers, and in situ engineering disturbance, it is extremely urgent to reveal the biogeochemical mechanism of DOM in sediment cores. To clarify the distribution pattern of DOM in pore water of sediments in eutrophic lakes, we collected the sediment core samples from the accumulated area, dredged area, and central area of western Lake Taihu, China. The optical characterizations of DOM in sediment pore water were compared by spectral indicators and parallel factor (PARAFAC) analysis. The diffusive fluxes of DOC at the water-sediment interface in the three areas were calculated by Fick's first law. These results can provide effective information for the vertical dynamics of DOM, the current potential efflux in sediments, and their links with lacustrine carbon cycling affected by human disturbance and natural environmental conditions in large shallow eutrophic lakes.

The sediment cores from three study sites AA, DA, and CA were sampled in April 2018 by using a gravity sediment sampler. According to the sampling and analysis methods of previous studies (Chen et al. 2017; Zhao et al. 2021), three parallel samples of sediment cores were collected from each study site. Each sample of the sediment core was sliced into layers of 1, 2, and 3 cm thickness, starting from the water-sediment interface. During the slicing process of sediment cores, a synchronous and accurate layering was maintained. All sliced sediment samples were stored in a sample box with an ice pack and transferred to the laboratory within 3 h. In the laboratory, the sediment samples were centrifuged for 10 min at 3,500 rpm to extract pore water samples. The pore water samples were filtered through the pre-burned (450 °C) Whatman glass microporous filter (aperture, 0.7 μm), and then, completely analyzed within 48 h. During the pretreatment process of sediment cores, we maintained synchronous and accurate layering. We measured the parameters of extracted pore water samples in each sliced layer of the sediment cores.

DOC and DIC concentrations from filtered pore water were directly detected using a carbon–nitrogen analyzer (multi N/C 3100; AnalytikJena, Jena, Germany) based on the non-purgeable organic carbon (NPOC) assay method. Fe (II) was detected using the Ferrozine colorimetrical method when total Fe ions was obtained by adding 10% hydroxylamine hydrochloride in order to reduce Fe³⁺ to Fe²⁺ for colorimetry (Lovley & Phillips 1988; Zhao et al. 2021).

Based on Milli-Q water that was used as a control, absorption spectra of pore water samples were scanned from 200 to 800 nm using a UV–Vis spectrophotometer (UV-3600, Shimadzu Inc., Japan) with a 1 cm quartz cuvette at room temperature. Absorbance data were used to calculate the absorption coefficient (a254) (Blough & Del Vecchio 2002; Chen et al. 2017), specific UV absorbance (SUVA₂₅₄) (Weishaar et al. 2003; Zhang et al. 2021), and the spectral slope (S_275-295 and S_R) (Helms et al. 2008).

Fluorescence EEMs of pore water samples were scanned using the PerkinElmer LS50B fluorescence spectrometer (Perkin Elmer Instruments Co., Ltd, USA) with a 1 cm quartz cuvette. The excitation wavelength (Ex) was set at 200–500 nm in steps of 5 nm. The emission wavelength (Em) ranged from 250 to 600 nm with a 0.5 nm interval and a scanning speed of 1,200 nm min⁻¹. Water Raman scatter peaks and Rayleigh scatter peaks were eliminated based on the corrected methods of previous studies (Stedmon & Bro 2008; Li et al. 2021). The EEM results were normalized as Raman units (R.U.) using the integrated area of the Raman peak of daily measured Milli-Q water excited at 350 nm (Lawaetz & Stedmon 2009). The EEM results can be used to calculate the fluorescence index (FI) (McKnight et al. 2001), the freshness index (FrI) (Hansen et al. 2016), the biological index (BIX) (Huguet et al. 2009), and the humification index (HIX) (Huguet et al. 2009) in sediment pore water of three typical areas.

The mean value and standard deviation of analysis parameters from the pore water samples of three parallel sediment cores at the same sediment depth were calculated as the final results (i.e., mean ± SD) at this sediment depth of the study site. The locations of the study area and sampling sites were plotted using ArcGIS 10.2 (Esri Inc., Redlands, CA, USA) and Coreldraw X7 (Corel Corporation, Canada). The comparison charts of data in different sampling sites and correlation analysis were analyzed and plotted using Origin 2021 (OriginLab Corporation, USA) and Microsoft Office Excel 2021 (Microsoft Corp, Redmond, WA, USA). The PARAFAC model was performed using MATLAB R2016a (MathWorks Inc., Natick, MA, USA) with the DOMFluor toolbox (Stedmon & Bro 2008). One-way analysis of variance (ANOVA) was performed to analyze differences in various characteristics using SPSS (version 22.0; IBM, Armonk, NY, USA).

Benthic fluxes of pore water DOC in situ sediment cores of three typical areas were estimated using Fick's first law (Table 1). According to the previous studies about MW of DOC in Taihu Lake (Huang et al. 2009), the estimations assumed an average pore water MW of 500–10,000 Da in this study. Based on this range of MW, the diffusion coefficients were calculated as 1.4–4.6 × 10–6 cm²·s⁻¹ through the log–log relationship between diffusion coefficients and molecular weights for various organic compounds (at 25 °C in distilled water) (Burdige et al. 1992). The estimated ranges of DOC benthic effluxes in algae accumulated, dredged, and central lake areas were −458.2 to −139.4 mg·m⁻²·d⁻¹, 9.5 to 31.2 mg·m⁻²·d⁻¹, and 14.6 to 48.0 mg·m⁻²·d⁻¹, respectively. The negative sign of the benthic efflux value in the algae accumulated area indicated that DOC diffused from overlying water to the sediment. Therefore, the diffusion direction of DOC at the sediment-water interface of the algae accumulated area was different from the other two areas.

In the algae accumulated area, many algae scums gathered in lake water or gradually settled into the sediment. This led to strong DOC release into overlying water via microbial degradation (Li et al. 2018; Ma et al. 2022). Accordingly, when the DOC concentrations of overlying water were higher than sediment pore water, it started to diffuse from overlying water to pore water through the sediment-water interface, which further resulted in a negative benthic efflux (Table 1). However, in the other two areas, POM including aquatic plants, algae, terrestrial, and dead aquatic animal debris stored in the sediment were turned into DOC by microbes, leading to higher DOC concentrations in sediment pore water compared to overlying water. Such differences indicated the opposite DOC diffusion direction of the two areas to the algae accumulated area at the sediment-water interface, which highlighted that sediments serve as DOC sources to the overlying water column in Lake Taihu (Li et al. 2021; Duan et al. 2022; Liu et al. 2023). After dredging, the long-term deposited sediment with abundant OM was removed and the current sediment has been primarily built up since dredging, compared with the central area (Zhang et al. 2010; Chen et al. 2021). Therefore, the benthic flux of pore water DOC in the dredged area was lower than in the central area.

It is noteworthy that DOC benthic fluxes in this study were comparable with those of lake and reservoir environments in previous studies, especially the ones whose spatial difference distributions were affected by various natural factors or human activities. Combining with the DOC diffusing flux variations of our previous microcosm experiment (Li et al. 2018), the DOC benthic flux of the algae accumulated area further suggested that the released DOC during algae decomposition alter carbon balance at the water-sediment interface and the carbon budget of overlying water and pore water in eutrophic lake ecosystems. This is further evidenced by other inland water ecosystems with lower trophic levels than Taihu Lake, which displayed the same diffusion direction of DOC in the dredged area and the central area, but higher benthic fluxes of inland water ecosystems, such as 51 ± 101 mg·m⁻²·d⁻¹ in Uiam Lake (Yang et al. 2014), 27–114 mg·m⁻²·d⁻¹ (Chen et al. 2017), and 88.3 mg·m⁻²·d⁻¹ in a nascent river-type lake (Niu et al. 2021). These benthic flux gaps are probably due to the different calculated processes or applying methods, for example, the uncertainty of diffusion coefficients due to variations of DOC molecular masses. Another possible reason is that DOC produced by the decomposition of alga bloom and aquatic plants enters overlying water, which further reduces the concentration gradient of DOC at the sediment-water interface (Li et al. 2018, 2021; Zhang et al. 2023).

During the decomposition process of algae scums in the algae accumulated area, abundant dissolved organic compounds were released into overlying water and then gradually transferred to the sediment profile (Li et al. 2018; Wang et al. 2020), resulting in higher concentrations of DOC at different depths than those of the other two areas. Fe (II) concentrations and Fe (II)/TFe of pore water can effectively indicate the anaerobic environment and the iron reduction process in sediment (Zhao et al. 2021). According to area differences and vertical gradient distribution of Fe (II) and Fe (II)/TFe, the algae accumulated area displayed a more reducing environment than those in the other two studied areas, because the dense reed barrier and the continuous micro/anaerobic environment caused by the accumulation of algae scums were conducive to the reduction of iron (Chen et al. 2016; Zhao et al. 2021). Meanwhile, a stable anaerobic environment was rare in the dredged area and the open water area due to dredging disturbance and wind waves, thus inhibiting iron reduction. From this perspective, the dynamic of the anaerobic environment clearly affected the migration and transformation between DOC and DIC in sediments (Li et al. 2018; Lu et al. 2022). In the algae accumulated area, the negative correlation between DIC and Fe (II) (r = −0.70, p < 0.01) suggested that the anaerobic environment is unfavorable for the transformation of DOC to DIC (i.e., and ⁠) and it probably promoted anaerobic methanogenesis (Figure 8(a); Tang et al. 2021; Zhang et al. 2022). Meanwhile, in the algae accumulated area, a higher DOC content in the deeper sediment (below 18 cm from the sediment-water interface) probably resulted from the accumulation of refractory DOC or DOM desorbed from reductive oxides under a hypoxic condition (Li et al. 2021).

The absorption coefficient a254 can effectively estimate the DOM concentration in aquatic ecosystems (Chen et al. 2017; Zhang et al. 2021), which was confirmed by the positive relationship (AA: r = 0.92, p < 0.01; DA: r = 0.95, p < 0.01; CA: r = 0.54, p < 0.01) (Figure 8) and the similarly vertical distribution between DOC and a254 in this study (Figures 2(a) and 4(a)). Moreover, the specific ultraviolet absorbance SUVA₂₅₄ could be used as a useful proxy for DOM aromatic content (Weishaar et al. 2003). The differences in SUVA₂₅₄ values between surface and deep sediments in the dredged area and the algae accumulated area indicated that the DOM aromatic contents were affected by the deposition and decomposition of algae-derived OM. In the two areas, the surface sediments were composed of recently settled and preferentially decomposed algae scums with lower aromaticity (SUVA₂₅₄) and a relatively high proportion of aromatic terrestrial DOM in the order of sediments due to the degradation of algae-derived components via bacterial activities (Li et al. 2022a, 2022b; Bao et al. 2023). Meanwhile, the SUVA₂₅₄ values first rose and then gradually decreased with depth, suggesting that the remaining DOM in deep sediments slowly decomposed. By contrast, the dredging work re-disturbed the vertical gradient characteristics of DOM, resulting in a big fluctuating pattern of SUVA₂₅₄ with surface sediment depths (Figure 4(b)). The lower SUVA₂₅₄ values at most depths in the central area than the other two areas further indicated that the lower aromaticity compounds associated with autochthonous DOM gradually migrated and accumulated through the sediment core (Zhang et al. 2021). In the algae accumulated area, the smaller vertical variation of S_275-295 and S_R than the other two demonstrated that the high MW of remaining refectory algae-derived and terrestrial dissolved compounds migrated and transformed slowly with sediment depths when the low MW of algae scums decomposed (Du et al. 2022a, 2022b; Li et al. 2022a, 2022b). Additionally, the maximum S_R values that occurred at 3 cm depth in the surface sediment probably resulted from the multi-mixed inputs of algae-derived DOC in overlying water and autochthonous DOM produced by microbial degradation of the sediment, which disturbed the stable and barely changed distribution of the remaining recalcitrant DOM along the sediment core (Zhang et al. 2023). In the central area far away from inflow rivers, the S_R values of the entire sediment core were close to or exceeded 1.0, which implied that the influences of autochthonous sources were more related to phytoplankton activities and microbial function on the migration and accumulation of DOM with sediment depths, compared to terrestrial inputs (Helms et al. 2008; Hansen et al. 2016; Zhang et al. 2023). This similar indication to the central area occurred in the dredged area although the dredging work caused a bigger fluctuation of higher S_R values in the sediment core. Particularly, the dredged sediment containing less microbial contents and species, had a deficiency of microbial degradation efficiency (Zhao et al. 2021), which was beneficial for the migration and accumulation of algae-derived labile DOM along the sediment depths.

In three study areas, a higher C1% and a lower C4% at surface sediments than deep sediments indicated that the terrestrial humic-like DOM gradually accumulated in the sediment core. Meanwhile, the autochthonous production of DOM is mainly concentrated in the surface sediment due to the high biodegradation of free and bound protein-like components produced by algae-derived DOM in eutrophic lakes (Li et al. 2018, 2022a, 2022b). The negative correlation between C2% and C4% in the algae accumulated area and the central area (AA: r = −0.89, p < 0.01; CA: r = −0.75, p < 0.01) proved that microbial activity had an important influence on the biodegradation and transformation of algae-derived DOM, causing the decreasing distribution of C4% (Figure 8(a) and 8(c)). In contrast, the dredged area did not show any significant correlation between C2% and C4%, suggesting that the dredging work re-disturbed the transformation and accumulation of DOM compounds in the sediment core (Figure 8(b)). The C3% of three study areas presented strong fluctuations with depths due to the complicated fractions and components of DOM related to agricultural humic-like substances, whose behaviors of degradation and transformation were uncertain in the lake ecosystem (Osburn et al. 2012; Gao et al. 2017; Zhou et al. 2018). Higher FI values in the algae accumulated area indicated that the accumulated amount of algae-derived DOM was more than the other two areas because the biomass of algae scums deposited and accumulated in overlying water and surface sediment of the accumulated area were much higher than the other two areas. Although some labile algae-derived DOM degraded and further transformed into DIC, CO₂, and CH₄, the remaining refectory algae-derived DOM maintained a higher total amount level through the entire sediment than the other two areas (Li et al. 2018). According to Frl and BIX differences of surface sediments, the relative proportion of autochthonous DOM derived from algae scums was higher in the other two areas than the accumulated area, which was affected by terrestrial inputs from inflow rivers. The recent accumulated algae scums in surface sediments caused higher Frl and BIX than the deep sediments in the study areas, as well as the corresponding decreasing humification degrees (HIX) with sediment depths. In the three areas, HIX presented a positive correlation with C1% but a negative relationship with C4% (Figure 8), indicating that the terrestrial DOM and remaining recalcitrant OM after the degradation of algae-derived labile OM contributed a higher humification for sediment DOM. Meanwhile, the higher HIX in the sediment core of the algae accumulated area than in the other two areas implied the deposition, accumulation, and transformation of massive algae scums in eutrophic lakes, which probably promoted the humification degree of sediments (Wang et al. 2020; Du et al. 2022a, 2022b).

In this study, the vertical distribution of DOC, inorganic Fe, absorption properties, and fluorescent composition in pore water of sediment cores were investigated to further unravel the differences of DOM sources, DOC diffusion at the sediment-water interface, as well as DOM transformation, accumulation, and humification from the algae accumulated, dredged, and central lake areas. Results showed that terrestrial inputs had a significant influence on the DOM components of pore water in the algae accumulated area while autochthonous sources contributed to DOM accumulation in sediment pore water in the central lake area. The dredging work disturbed the pre-dredging distribution patterns of DOM compounds in sediment cores. The diffusion directions of dredged and central lake areas at the sediment-water interface were opposite to the algae accumulated area because of the strong release of DOC into overlying water during the decomposition of massive algae scums in lake water. Our results further suggested that the promoted humification degree of sediments resulted from the deposition, accumulation, and transformation of algae-derived DOM in eutrophic lakes.

This research was supported by the National Natural Science Foundation of China (42107280), the Scientific Research Projects of Colleges and Universities in Anhui Province (2022AH040211 and 2022AH040208), the Postdoctoral Scientific Research Foundation of Suzhou University (2021bsh003), and the Doctoral Scientific Research Foundation of Suzhou University (2019jb26, 2020BS021).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.