Organic Carbon in the Bottom Sediments of Lake Baikal: Geochemical Processes of Burial and Balance Values

Abstract

This is the first study of dissolved organic matter (DOM) at the Lake Baikal water-bottom interface. High-resolution profiles of dissolved organic carbon (DOC) were obtained simultaneously with dissolved inorganic carbon (DIC), total dissolved carbon, cations (Na⁺, K⁺, Ca²⁺, Mg²⁺, Fe²⁺, and Mn²⁺), and anions (HCO₃⁻, Cl⁻, NO₃⁻, and SO₄²⁻) in the pore water of Lake Baikal deepwater oxidized sediments. We evaluated the DOC fluxes quantitatively and qualitatively. They changed their direction twice under different redox conditions in the sediments (at the redox interfaces). The study revealed that the mobilization of DOC in anoxic sediments was closely related to the reductive dissolution of Fe(III) minerals, and the oxidized surface lake sediments represented an effective DOC trap binding DOC to ferric minerals. Redox conditions appeared to be the main regulator of the DOC exchange. Oxygen conditions led to the uptake of DOC by sediments (31–78 mmol C m⁻² yr⁻¹), i.e., the Lake Baikal sediments are a sink of DOC. The DOC flux was approximately 25–35% of the carbon flux at the sediment–water interface. The results of this study allow for a better understanding of the nature and properties of DOC in freshwater ecosystems and compensate for the underestimation of DOC in the internal carbon cycle of the lake.

1. Introduction

Lakes play an important role in the global carbon cycle and carbon balance. They can store large amounts of organic carbon (OC) in sediments and thus exclude it from active turnover, representing an important global carbon stock. The amounts of buried carbon are comparable to those in the ocean [1]. Carbon buried in bottom sediments can ultimately immobilize or mineralize into carbon dioxide and/or methane and diffuse out of the sediments [2]. Moreover, organic carbon in the solid part of the sediments can be released after mineralization into the pore water and diffuse out of the sediment in the form of dissolved organic carbon (DOC) [3]. Disregard for this part leads to an undercount of organic carbon in the overall balance of the carbon cycle.

The trend of increasing dissolved organic carbon in surface waters contributes to the study of DOC processes [4]. A deep insight into internal cyclic processes and interactions is important for determining the dynamics of DOC concentrations. Particulate organic matter deposition is a well-known one-way process directed from water to sediment. On the other hand, DOC flux is a two-way process carried out by diffusion, advection, and resuspension of the sediment. Thus, bottom sediments can be either a source or a sink for DOC. Microbial decomposition of particulate organic matter usually leads to an increased concentration of DOC in the pore water of the bottom sediments [5]. However, according to the data [6,7,8,9,10,11], not only metabolic processes but also the interaction with mineral surfaces influence the concentration of DOC in the presence of mineral particles. Recent studies of the lake sediments have indicated that DOC flux is associated with adsorption and chelation processes where organic molecules are “glued” by iron oxide nanophases. These processes are reversible and controlled by the redox state of iron [7,8,12,13]. Similar studies have not yet been conducted at Lake Baikal. DOC was studied two decades ago and only in the water column of the lake [14].

Lake Baikal, located in East Siberia (Russia), is the world’s largest freshwater lake, containing 20% of the global freshwater reserves (23,015 km³) and being the deepest (1642 m) and oldest (20–50 million years) lake on Earth. Its sediments are up to 7.5 km thick, the oldest of which are of Oligocene age [15]. The lake is morphologically divided into three deep-water basins (southern, central, and northern) that are separated from each other by two underwater elevations: the Buguldeika-Selenga Isthmus and Akademichesky (Academic) Ridge [16]. A compensated water exchange takes place between neighboring basins; horizontal water exchange caused by cyclonic macro-circulation is observed within each basin [17]. Significant flat basin plains allow easy accumulation of the pelagic sediments. Efficient vertical mixing (twice a year) renews near-bottom waters and promotes high oxygen concentrations throughout the lake (9.6–12.8 mg L⁻¹), including near-bottom water [18]. Low sedimentation rates (0.5–0.9 mm yr⁻¹) and low productivity of cold oligotrophic lakes contribute to oxygen penetration into the bottom sediments (over 5 cm), leading to the oxidative stage of diagenesis [19,20]. Surface sediments are oxidized almost throughout the entire lake bed [21].

The formation of iron and manganese layers within the Baikal sediments is a unique feature. Granina et al. and Maerki et al. suggested that the accumulated layers resulted from a complex interplay of the Corg mass accumulation rate (gC m⁻² yr⁻¹), the sedimentation rate (mm yr⁻¹), the oxygen supply of the bottom water, and the sequence of diagenetic redox reactions as well as molecular diffusion of substances within the sediment pore water [19,22]. Och et al. hypothesized a cycle characterized by the dynamic growth of Fe and Mn oxide layers immediately underneath the maximum O₂ penetration depth, a reductive dissolution of the Fe/Mn oxide accumulation, and the subsequent initiation of a new dynamic Fe/Mn layer above [20]. Och et al. and Torres et al. also suggested that the dissolution of Fe/Mn oxide was coupled to the anaerobic oxidation of CH₄ (AOM) either by the reduction in sulfate and the subsequent generation of Fe(II) by S(II) oxidation or directly coupled to Fe reduction [20,23]. The study of Och et al. also demonstrated that the Fe/Mn layers were associated with the diagenetic redistribution of REE [24], the mobility of which was influenced by the adsorption and desorption dynamics in the process of dissolution and reduction in the Fe/Mn oxides. Similar processes were confirmed in the study of uranium [25]. Significant U enrichments prevailed within the top oxygenated layer of the sediment due to U adsorption to and/or co-precipitation with Fe-oxides. When Fe-oxides and, to a lesser extent, Mn-oxides were reductively dissolved, they released U into the pore water, leading to peak dissolved U concentrations in the anoxic sediment.

The first data on DOC in pore water obtained during two expeditions to Baikal in September 1988 and 1991 (more than thirty years ago) were published in 2020 [26]. The authors presented the data on carbon deposition, remineralization, stoichiometry, and recycling in sediments and noted discrepancies in estimates through recycling flux and through diagenetic mass balance. The problems were probably because the authors did not take into account the Fe/Mn cycling of the oxidative stage of diagenesis in the Lake Baikal sediments.

Considering the abovementioned issues, it became necessary to investigate the processes and interactions of Fe/Mn cycling in relation to DOC, which have not yet been conducted at Lake Baikal.

In this study, we present a conceptual model of organic carbon burial in Lake Baikal sediments. Based on high-resolution profiles of DOC (together with DIC, cations, and anions) in pore water from Lake Baikal deepwater oxidized sediments and following the previous studies [19,20,23,24,25,26], we examine: (i) the sorption properties of different Fe(III) phases, the absorption of DOC, cations, and anions, and the effect of calcium; (ii) the dependence of the Corg cycle on the geochemical Fe cycle and influence of the sediments redox conditions; (iii) the direction, magnitude, and regulation of the benthic DOC flux in Lake Baikal sediments; and (iv) the participation of DOC in the overall balance of the carbon cycle.

2. Materials and Methods

2.1. Sampling

The bottom sediments were sampled with a benthic corer (diameter 100 mm and length 1 m) in the pelagic zone of the southern (depth 1476 m) and northern (depth 925 m) basins of Lake Baikal (Figure 1) during the 2020 expedition (

Table 1. Sampling locations and positions, dates, and depths during the Lake Baikal expedition.

Location	Station	Core	Date	Position	Lake Depth, m	Core Length, cm	Number of Samples
Southern basin	A	Ver 20 01 St13BGC1	24 June 2020	51°42′10.4″ N 104°58′47.39′′ E	1476	55	19 *
Northern basin	B	Ver 20 01 St6BGC1	19 June 2020	54°14′16.99″ N 108°48′16.38′′ E	925	50	33 *

Note: * Each analysis was carried out in triplicate.

).

2.2. Pore Waters and Benthic Water Chemical Analysis

Pore waters were obtained on board the RV “Vereshchagin” immediately after core sampling by centrifugation of sediments for 20 min at 8000 rpm, then 10 min at 14,000 rpm, followed by 0.20 µm filtration (cellulose acetate, Vladisart, Vladimir, Russia).

The concentration of anions (HCO₃⁻, SO₄²⁻, NO₃⁻, and Cl⁻) in pore waters was measured on board the RV immediately after receiving solutions by liquid chromatography on a Milichrom-2A chromatograph (Ekonova, Novosibirsk, Russia) according to the method in [27] with a relative error of 5–7% (for chlorides up to 10%). Pore water cations (Na⁺, K⁺, Ca²⁺, Mg²⁺, Fe²⁺, and Mn²⁺) were determined in the laboratory shortly after the expedition by atomic absorption and flame emission methods (relative error of 2–3%). Samples were stored in a refrigerator at 4 °C until analysis. A more detailed description of the methods is given in [28].

Concentrations of dissolved organic carbon, dissolved inorganic carbon, and total dissolved carbon in pore waters were measured using a Vario TOC Cube high-temperature carbon analyzer with an infrared detector (Elementar Analysensysteme GmbH, Hanau, Germany). The standard deviation of the determinations did not exceed 0.01. Potassium biphthalate (Kanto Chemical Co. Ltd., Tokyo, Japan) was used as a standard. Before the analysis, samples were frozen immediately after solution filtration and stored at −20 °C.

The determination of oxygen content in benthic water was performed by the standard Winkler method. Water was decanted from the benthos tube immediately after core extraction.

2.3. Redox Potential (Eh) of Sediments

Redox potential (Eh) was determined in wet sediments using a pH meter (WTW portable pH meter ProfiLine pH 3310 WTW, Weilheim, Germany) with a separate ORP electrode (Eh accuracy ± 0.3 mV). Zobell solution (YSI Incorporated, Yellow Springs, OH, USA) was used for calibrating ORP/redox.

2.4. Sediment: OC, Iron, and Manganese Extraction, Analysis, and Calculations

To determine the amount of reactive iron and manganese and organic carbon bound to iron and manganese (oxyhydroxides), we used the method described in detail by [7,11,29]. A total of 15 mL of a solution containing 0.27 M trisodium citrate (Na₃C₆H₅O₇ × H₂O) and 0.11 M sodium bicarbonate (NaHCO₃) were added to 0.25 g of the dried and homogenized sediment sample, then stirred over and heated to 80 °C in a water bath. Thereafter, we added 0.25 g of 0.1 M sodium dithionite (Na₂S₂O₄), the temperature was maintained at 80 °C, and the tubes were shaken every five minutes [11]. After 15 min, the samples were centrifuged for 10 min at 3360× g. The solution was decanted, and 200 µL of HCl was added to prevent Fe(III) precipitation. The remaining sediment was washed and dried three times. The losses of organic carbon, which were not associated with iron oxides, were counted according to an identical scheme but without complexing and reducing agents. After the treatment, all samples were weighed to estimate the mass loss.

Dissolved iron and manganese in the supernatant liquid were analyzed using an atomic absorption spectrometer (ContrAA800 Analytic Jena, Jena, Germany) (relative error 2%). The OC concentration in the sediments before and after reduction and control was determined using a Vario TOC Cube carbon analyzer with an infrared detector (Elementar Analysensysteme GmbH, Hanau, Germany). The relative error of the method was 2.3%.

The mass balance, taking into account dissolution losses, was calculated according to [29]:

$$\mathrm{O}\mathrm{C}-\mathrm{F}\mathrm{e}\mathrm{R}\left(\mathrm{\%}\right)=100\times \left(\mathrm{T}\mathrm{O}\mathrm{C}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}-\mathrm{T}\mathrm{O}\mathrm{C}\mathrm{a}\mathrm{f}\mathrm{t}\mathrm{e}\mathrm{r}\right)\times \left(\mathrm{M}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}/\mathrm{M}\mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}\right)/\left(\mathrm{T}\mathrm{O}\mathrm{C}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)$$

(1)

where OC − FeR(%) is the percentage of organic carbon bound to reactive iron; TOC_control is the total organic carbon content after the control extraction; TOC_after is the total organic carbon content after the reduction extraction; TOC_initial is the total organic carbon content in the sediment; M_control is the mass before the control extraction; and M_sample is the mass before the reduction extraction.

A similar technique was used to determine the amount of reactive manganese and organic carbon associated with it.

2.5. Calculations of Diffusion Fluxes of Dissolved Organic Carbon

Diffusion fluxes of dissolved organic carbon were calculated by chemical gradients of pore water concentration profiles using Fick’s first law [30], taking into account porosity and tortuosity [20,23,31]:

$$\mathrm{J}\mathrm{s}\mathrm{e}\mathrm{d}=-\mathsf{\phi }\times \mathrm{D}\mathrm{s}\mathrm{e}\mathrm{d}\times \partial \mathrm{C}/\partial \mathrm{x}$$

(2)

$$\mathrm{D}\mathrm{s}\mathrm{e}\mathrm{d}=\mathrm{D}\mathrm{o}/(\mathsf{\phi }\times \mathrm{F})$$

$$\mathrm{F}=1.02\times {\mathsf{\phi }}^{-1.81}$$

where J is the diffusion flux; φ is the porosity (counted according to [20] in similar sediments); F is the coefficient correcting for porosity and tortuosity of sediment [31]; $\partial$C/$\partial$x is the concentration gradient; and D$\mathrm{o}$ is the diffusion coefficient in pore water at 5 °C [32] (molecular diffusion for DOC was estimated regarding the molecular weight as ~1000, being consistent with measurements in the lakes [3,6,13,33]; temperature dependence of DOC diffusivities was assumed to follow the Stokes-Einstein relation for slow diffusing constituents [32]. Do = 1.6 × 10⁻⁶ cm² s ⁻¹ [13,26,33]).

3. Results

3.1. Lithology of Bottom Sediments

The sediments sampled at the deepwater stations of the southern and northern basins of Lake Baikal were typical of these locations [20,23,24,25]. The sediments contained gray biogenic–terrigenous diatom silt and were oxidized from the surface to a reddish-brown ochre color (5 cm thick in the southern basin and 19 cm thick in the northern basin) (Figure 1). Fe-Mn crusts were found at the interface of oxidized and reduced sediments. The sediments resulted from continuous sedimentation without any traces of bioturbation. There was a sandy interlayer at the 14 cm layer of the core from station B (northern basin).

3.2. Oxygen

The oxygen concentration in the bottom water was 12.4 mg O₂ L⁻¹ (0.39 mmol L⁻¹) in the southern basin and 13.1 mg O₂ L⁻¹ (0.42 mmol L⁻¹) in the northern basin, which was typical and corresponded to [20]. Therefore, this study discusses the depth of O₂ penetration into sediment (Figure 2), which was measured and discussed for the same areas in [20], namely, 2 cm in the southern basin and 6 cm in the northern basin. The calculations [20] yielded O₂ fluxes of 5.6 mmol m⁻² d⁻¹ (station A) and 4.0 mmol m⁻² d⁻¹ (station E).

3.3. Redox Values

In both the southern and northern basins of Lake Baikal, the positive values of the redox potential (Eh) at the water-bottom interface of +159–+170 (Figure 2 and Figure 3) dramatically increased in the first subsurface centimeters of sediments (O₂ penetration depth) to +221–+272 and then decreased, passing into negative values at the redox boundary of oxidized and reduced sediments. With depth, the decrease continued with a smaller gradient, reaching values of −195–−205 per 45 cm of sediment.

3.4. Pore Water Chemical Composition

The distribution profiles of pore water components in the sediments were similar in the southern and northern basins of Lake Baikal, but they were compressed or extended in depth, depending on the thickness of the oxidized layer (Figure 2). The concentrations of Cl⁻, K⁺, Mg²⁺, and Na⁺ in pore water throughout the depth of the sediment profile varied insignificantly. The concentrations of Fe²⁺ and Mn²⁺ in pore waters increased and reached their maximum values, forming a peak with a maximum at a depth just below the interface between the oxidized and reduced sedimentary zones (6 cm in the southern Baikal basin and 20 cm in the northern Baikal basin). Significant similarity was found in the distribution of Ca²⁺ and dissolved carbon (organic, inorganic, and total) in pore waters, with correlation coefficients of 0.75–0.92. Meanwhile, a high correlation between concentrations of HCO₃⁻ (actually DIC) and Ca²⁺ was previously observed [34]. The concentrations of Ca²⁺, DIC, and DOC in pore waters decreased significantly to a depth of 2 cm in the southern basin and 6 cm in the northern basin (O₂ penetration depth). Then, it increased, forming a peak with a maximum (the same as Fe²⁺ and Mn²⁺) at a depth below the interface between the oxidized and reduced sediment zones (6 cm in the southern basin and 20 cm in the northern basin). Thereafter, the concentration decreased in the reduced sediments and did not change significantly throughout the core.

SO₄²⁻ concentrations did not exceed 80 μmol L⁻¹ in the pore waters from both stations (the southern and northern basins of Lake Baikal). SO₄²⁻ concentrations tended to increase slightly within the oxygenated part of sediments and decrease below the O₂ penetration depth. Concentration peaks also formed below the interface of the oxidized and reduced sediment zones (Figure 2). NO₃⁻ concentrations did not exceed 17 μmol L⁻¹ in the pore waters from both stations (the southern and northern basins of Lake Baikal). A slight concentration peak formed near the O₂ penetration depth interface (Figure 2).

3.5. Reactive Iron and Organic Carbon Associated with Iron

As far as the distribution profiles of pore water components in the sediments of the southern and northern basins of Lake Baikal had a similar nature and orientation relative to the thickness of the oxidized layer, the procedure of reducing iron dissolution was performed only for the core from station B as the most informative because of the large thickness of the oxidized layer and profiles of higher resolution.

Reductive dissolution of iron with dithionite allowed us to determine the entire pool of reactive iron oxides, FeR [11], capable of binding, the amount of organic carbon bound to iron oxides (OC-FeR), and its proportion in the total amount of organic carbon (fOC-FeR). The results obtained are shown in Figure 3.

In the oxidized sedimentary layer, the amount of reactive iron (FeR) increased from the surface to 6 cm (oxygen penetration depth), reaching 3.5%, then decreased smoothly (a sharp minimum corresponded to the sandy interlayer). The maximum FeR content of 5.9% referred to the Fe-Mn crust at the interface of the oxidized and reduced sedimentary layers.

Total organic carbon (TOC) in the core ranged from 2.8% in the upper layer to 0.7% in the deeper layers (0.4% corresponded to the sandy interlayer). There was a clear tendency for a TOC decrease in the upper oxidized sedimentary layer (Figure 3), and the concentrations were more stable, varying about 0.6–1.1% in the lower reduced part.

The analysis showed that a significant proportion of TOC in the upper oxidized layer of the sediment (Figure 3) was correlated with reactive iron. The proportion of organic carbon associated with iron oxides in the total organic carbon (fOC-FeR) increased from 18% at the surface to 54% at 6 cm (oxygen penetration depth) and then decreased gradually to 3% in the reduced part of the sediments. The molar ratio of iron-associated organic carbon to reactive iron (OC:Fe) showed low values close to 1 (Figure 3).

3.6. Reactive Manganese and Organic Carbon Associated with Manganese

A reductive manganese dissolution procedure involving dithionite revealed that the proportion of reactive manganese capable of binding carbon did not exceed 0.02% throughout the sediment cores.

4. Discussion

Numerous biogeochemical processes control the DOC concentration in the pore water of the bottom sediments [8]. The detected Z-shaped profile of DOC indicates different processes in the uppermost surface layer and deeper sedimentary layers.

It is well known that the bottom water of Lake Baikal contains a large amount of oxygen throughout the year (according to our data, the southern basin contains 12.4 mg L⁻¹ of O₂, and the northern basin contains 13.1 mg L⁻¹), which also penetrates the bottom sediments. According to [20], O₂ penetration depth at the same stations was 2 cm in the southern basin and 5 cm in the northern basin, so the sharp decrease in the DOC concentration in the uppermost layer of sediments may be due to aerobic decomposition of organic matter. Indeed, some publications reported that in oxygen-containing sediments, dissolved organic matter (DOM) consisted of organic compounds of lower molecular weight as a consequence of OM degradation [6,33]. However, other studies [9,13,35] indicated the predominance of Fe-related processes. According to these data, coagulation and sorption of DOM on Fe(III) hydroxides in the oxygen surface layer (as the faster processes than degradation) can take out up to 50% of DOC from the solution, mainly with aromatic and high-molecular-weight compounds [35]. The strong relationship between iron hydroxides and DOC can prevent the decomposition of organic matter by microorganisms and contribute to its preservation.

4.1. Peculiarities and Influence of Sorption Properties of Iron Hydroxides. Features of the Absorption of Dissolved Organic Matter, Cations, and Anions

In the sediments of Lake Baikal, at the water-bottom interface, Fe(III) mainly represents Fe, about half of which has amorphous or weakly crystalline phases of Fe(III) [25]. Amorphous ferrihydrite (Fh) is a natural nanoparticle. The particles are extremely small in fresh deposition: they are within a few nanometers in diameter and have an increased surface charge; the reactive surface area is vast, about 650–1100 m² g⁻¹ [36,37]. Therefore, even small amounts of these nanoparticles can dominate the ion-binding properties. However, during senescence, the particles accumulate and merge with each other, reducing the surface area to 200–400 m² g⁻¹. Fh has a very high surface density of binding sites, especially in the single-coordinated (≡FeOH) groups. At the same time, two- and three-fold coordinated surface group nodes (≡Fe₂O and ≡Fe₃O, respectively) can also be formed, which can bind protons, resulting in a total surface charge with an uneven distribution. Owing to its emerging amphoteric property, ferrihydrite is capable of adsorbing both metal cations and oxy-anions [37]. Moreover, specific surface areas bind ions with different high and low affinities. Recent research [38] has identified a series of increasing affinities for Fh (log K) corresponding to an increase in the radius of the ions. The series was opposite to the order of the Hofmeister series observed for other Fe-(hydr)oxides: goethite and hematite.

The data obtained in the southern and northern basins (Figure 2) clearly reflect different changes in the concentration profiles of pore water ions from the upper oxidized layer of sediments. Cl⁻, K⁺, and Na⁺ do not almost change; Ca²⁺, HCO₃⁻ (DIC), and DOC change dramatically. The series of variations obtained from the gradients of changes in Cl⁻ < K⁺ < Na⁺ < Mg²⁺ < Ca²⁺ ≈ HCO₃⁻ (DIC) < DOC is in good agreement with the series of increases in the affinity to ferrihydrite, indicating its predominant role in adsorption among other Fe-(hydr)oxides of the surface sediments from Lake Baikal.

Previously, our experiments [34] demonstrated the significant absorption of ions on ferrihydrite. The treatment of Baikal water with the freshly prepared Fe(OH)₃ paste caused the complete removal of HCO₃⁻, Ca²⁺, and Mg²⁺ from the solution and decreased the concentrations of SO₄²⁻, Na⁺, and K⁺ by factors of 3, 1.3, and 1.3, respectively. The contact of Baikal water with the unwashed oxidized sediments decreased the concentrations of HCO₃⁻, Ca²⁺, and Mg²⁺ by a factor of 1.4, with a simultaneous reduction in pH from 7.8 to 6.6.

4.2. Features of the Absorption of Dissolved Organic Matter by Fe (Oxyhydr)oxides. The Effect of Calcium on the Absorption of DOC

According to recent data, dissolved organic matter has a very high affinity for adsorption by ferrihydrite [39]. Our data demonstrate a sharp two-fold decrease in the DOC concentrations in the upper oxidized layer of sediments (Figure 2). The adsorption of DOM on Fe (oxyhydr)oxides (especially on ferrihydrite) is known to be closely related to the molecular properties and causes fractionation [35,40,41,42]. Molecules with high molecular weight (MM) and unsaturated or oxygen-rich molecules, including condensed aromatic compounds, polyphenols, and carboxylic compounds, are selectively adsorbed by Fe (oxyhydr)oxides, while molecules with low molecular weight and molecules having low unsaturation or containing few oxygen-containing groups remain in solution. The equivalence of the double bond and the number of oxygen atoms play a major role here [40]. Additionally, amorphous Fh shows stronger bonds and fractionation than crystalline Fe (oxyhydr)oxides such as lepidocrocite and goethite [40]. Notably, during senescence, the transformation of ferrihydrite into other Fe (oxyhydr)oxides, DOM is desorbed into the solution and is no longer sorbed by the resulting lepidocrocite, goethite, and hematite [39,43]. The latter discovery is very important for understanding the processes during diagenesis of the lake surface sediments.

The presence of Ca ions in the solution has a great influence on the adsorption of DOC by ferrihydrite. For Ca²⁺/DOC ≥ 0.026, OM forms a large Ca-branched network within the Fe-DOM aggregates, in which Ca²⁺ is bound as a dimer to DOM carboxylic sites [44]. Competition between Ca and Fe for OM binding leads to an increase in the number of ferrihydrite-like nanoparticles and a greater availability of binding sites. The formation of such a micrometric grid significantly increases the total sorption capacity of the units. In the pore waters of the investigated stations, the Ca²⁺/DOC ratio of ≈ 0.4 suggests the formation of a micrometric grid and increased sorption capacity. A sharp two-fold decrease in the DOC concentrations in the upper oxidized layer of sediments both in the southern and northern basins (Figure 2) is accompanied by a similar decrease in Ca²⁺ concentrations, with correlation coefficients of 0.75 and 0.92, respectively.

4.3. Influence of Redox Conditions on the Sediments

The redox conditions of sediments determine the amount of reactive iron capable of binding (FeR) [45]. In the surface sediments of Lake Baikal, Fe/Mn-enriched layers are formed at the lower part of the oxygen penetration depth, moving upwards as the sediment grows [20]. At the lower interface of the oxidized layer, slow reductive dissolution of the buried lower Fe/Mn layers occurs, and the upward diffusing Fe²⁺ and Mn²⁺ are oxidized by O₂ near the interface of O₂ penetration depth. Our data indicate the maximum redox potential at the interface of oxygen penetration and the formation of fresh Fe (oxyhydr)oxides (FeR) with high binding capacity (Figure 3), especially ferrihydrite (according to [25]). Here, we also observe the maximum amount of reactive iron that binds the maximum amount (54%) of organic carbon, fOC-FeR (Figure 3). Moreover, there is a very good correlation (R² 0.72) between the Eh and fOC-FeR profiles and a very good correlation (R² 0.92) between the Eh profiles and the amount of reactive iron (Figure 4). Thus, our data confirms that the redox conditions of the sediments determine the amount of reactive iron (FeR) capable of binding and the amount of carbon associated with reactive iron (fOC-FeR).

The redox conditions have a pronounced effect on the binding processes. Below the oxygen depletion depth, a decrease in Eh promotes the reductive dissolution of Fe(III) to Fe(II) in sediments. The iron reduction consumes protons resulting in an increase in pH and general attenuation of OM sorption and, consequently, the release of sorbed DOM from the sediment solid phase [9,46]. Ultimately, we can observe synchronous peaks of Fe²⁺ and DOM slightly below the interface of oxidized and reduced sediments (Figure 2). Thus, the increase in DOM is due to its release resulting from iron reduction rather than a decrease in microbial mineralization of DOM. Balancing the processes of desorption and anaerobic degradation of OM determines further reductions in the concentration of DOM in the reduced sediments.

Noteworthy is the larger lower FeR peak (Figure 3) corresponding to the lower buried Fe/Mn layer, which demonstrates the minimum organic carbon fixation in the oxidized layer (8%). According to [25], crystalline phases of iron oxide, such as goethite and hematite, predominate here. The maturation and crystallization of FeR from fresher phases, such as ferrihydrite, to goethite/hematite reduces the surface area and reactivity, while DOC sorbed by ferrihydrite, according to [39,43], desorbs into solution and is no longer sorbed by the resulting goethite and hematite.

4.4. Factors Affecting the Stabilization of OC

An important factor influencing the stabilization of OC is adsorption or co-precipitation, a mechanism of binding FeR-OC, because co-precipitation shows greater conservation due to the low desorption of co-precipitated OC under changing redox conditions [7]. Chen and others [47] indicated the significance of the molar ratio characteristic where co-precipitation was determined for OC:FeR ≥ 2.8 and sorption was determined for OC:FeR ≤ 2.8. Our studies have demonstrated low values of molar ratios close to one (Figure 3), confirming that sorption is the dominant mechanism of carbon binding in the bottom sediments of Lake Baikal and that co-precipitation may have only a limited contribution.

Manganese can also potentially interact with organic substances [48]. However, our studies revealed the presence of reactive manganese two orders of magnitude lower than reactive iron (< 0.02%MnR), indicating that its role cannot be significant.

4.5. Conceptual Model of Organic Carbon Burial

Mineralization of organic matter in sediments near the water-bottom interface of Lake Baikal proceeds through a complex system of decomposition reactions combined with microbial-mediated oxidation. This is reflected in the consistent consumption of terminal electron acceptors such as O₂, NO₃⁻, Mn²⁺, Fe²⁺, and SO₄²⁻ with bursts of accumulation peaks during remineralization at the redox interfaces of the sediments (Figure 2). However, the quality and bioavailability of DOC to microorganisms under conditions activating sorption/desorption processes and interaction with Fe oxyhydroxides play a major role here. Taking into account the literature and our abovementioned data, we propose a basic scheme of organic carbon burial in the bottom sediments of Lake Baikal (Figure 5).

Thus, the process of burial of organic carbon at the water-bottom interface is a complex set of biogeochemical processes, including both microbial degradation of organic matter and adsorption/desorption on Fe hydroxides.

4.6. DOC Fluxes and Adjustment of Balance Values

Sedimentation of particulate organic matter is a one-way process directed from the water to the sediment. The flux of DOC is a two-way process where bottom sediments can be both the source and stock of DOC. Our studies have shown that adsorption processes at the water-bottom interface, which depend, in turn, on redox conditions, affect the direction and magnitude of the DOC flux. The calculated DOC fluxes (Table 2) at the water-bottom interface in the surface sediments of Lake Baikal change direction twice at the interfaces of redox potential, accounting for 78, −102, and 5.9 mmol C m⁻² yr⁻¹ in the southern basin of Lake Baikal and 31, −33, and 8.6 mmol C m⁻² yr⁻¹ in its northern basin. Thus, at the water-bottom interface, there is a positive DOC flux, i.e., the flux from water to sediment. Taking into account the large amount of oxygen at all depths of the lake, including bottom water, and the oxidized sediments covering almost the entire bottom of the lake, it can be argued that the sediments of Lake Baikal are a sink for DOM.

The DOC flux in this study is comparable both in direction and magnitude with previous studies [26], taking into account that [26] have calculated DOC flow under anoxic conditions. Our data is in good agreement both in direction and magnitude with the studies of deepwater marine sediments [49,50] and is lower by one or two orders of magnitude than the shallow water ones [3,4,8,13] (Table 3). These observations indicate that the DOC fluxes probably depend on the deposition rate and quality of OC, O₂ availability at the sediment surface, and aquatic system depth.

Table 2. Organic carbon fluxes at the water-bottom interface of Lake Baikal.

Areas	Zones in Sediments	Depth, cm	OC 1 Flux mmol m−2 yr−1	DOC 2 Flux mmol m−2 yr−1
Southern basin			220
	I	0.1–2		78
	II	2–6		−102
	III	6–45		5.9
Northern basin			125
	I	0–6		31
	II	6–21		−33
	III	21–45		8.6

Notes: ¹ Burial of organic carbon (particulate matter) according to sediment traps and O₂ consumption for mineralization at the water-bottom interface of Lake Baikal [51]; ² Negative values indicate upward fluxes: from the sediments to the bottom water or from deep sediments to the surface and vice versa.

Table 2. Organic carbon fluxes at the water-bottom interface of Lake Baikal.

Areas	Zones in Sediments	Depth, cm	OC 1 Flux mmol m−2 yr−1	DOC 2 Flux mmol m−2 yr−1
Southern basin			220
	I	0.1–2		78
	II	2–6		−102
	III	6–45		5.9
Northern basin			125
	I	0–6		31
	II	6–21		−33
	III	21–45		8.6

Notes: ¹ Burial of organic carbon (particulate matter) according to sediment traps and O₂ consumption for mineralization at the water-bottom interface of Lake Baikal [51]; ² Negative values indicate upward fluxes: from the sediments to the bottom water or from deep sediments to the surface and vice versa.

Table 3. Benthic DOC fluxes in aquatic systems under different redox conditions.

System	Oxygenation	Water Depth, m	DOC 1,2 Flux mmol m−2 d−1	References
Freshwater
Southern basin	oxic	1476	0.21	This study
	anoxic		−0.27	This study
Northern basin	oxic	925	0.09	This study
	anoxic		−0.09	This study
Southern basin	-	1415	−0.17	[26]
Northern basin	-	897	−0.05	[26]
Lake Erssjön	anoxic	4.5	−3.3	[13]
Sweden				[13]
Hassel reservoir	oxic	5	2.4	[4]
Germany	anoxic		−1.8	[4]
Uiam Lake	oxic	13	4.8	[8]
South Korea	anoxic		−5.1	[8]
Marine
Arabian Sea	anoxic	3190–4420	−0.06–−0.22	[50]
NE Atlantic	anoxic	4500–4800	−0.05–−0.12	[50]
Baltic Sea	oxic	231	0.02	[49]
	anoxic		−0.44	[49]
Long Island Sound	oxic	15	1.5	[3]
USA	anoxic		−5.2	[3]

Notes: ¹ The data are reduced to a single system due to the divergence of flow directions by different authors: positive flows are directed into the sediment, negative flows are directed from the sediment; ² The data are reduced to a single system of units.

Table 3. Benthic DOC fluxes in aquatic systems under different redox conditions.

System	Oxygenation	Water Depth, m	DOC 1,2 Flux mmol m−2 d−1	References
Freshwater
Southern basin	oxic	1476	0.21	This study
	anoxic		−0.27	This study
Northern basin	oxic	925	0.09	This study
	anoxic		−0.09	This study
Southern basin	-	1415	−0.17	[26]
Northern basin	-	897	−0.05	[26]
Lake Erssjön	anoxic	4.5	−3.3	[13]
Sweden				[13]
Hassel reservoir	oxic	5	2.4	[4]
Germany	anoxic		−1.8	[4]
Uiam Lake	oxic	13	4.8	[8]
South Korea	anoxic		−5.1	[8]
Marine
Arabian Sea	anoxic	3190–4420	−0.06–−0.22	[50]
NE Atlantic	anoxic	4500–4800	−0.05–−0.12	[50]
Baltic Sea	oxic	231	0.02	[49]
	anoxic		−0.44	[49]
Long Island Sound	oxic	15	1.5	[3]
USA	anoxic		−5.2	[3]

Notes: ¹ The data are reduced to a single system due to the divergence of flow directions by different authors: positive flows are directed into the sediment, negative flows are directed from the sediment; ² The data are reduced to a single system of units.

Our data, as well as the literature data (Table 3), indicate that sediments are a DOC source under anoxic conditions but a DOC sink under oxic conditions.

The divergence between the elevated negative DOC flux in the sediments from the southern basin and the positive DOC flux from the water into the sediment can be explained by the DOC consumption at the sediment surface. Our vertical resolution of pore water in the southern basin sediment does not elucidate these processes at the very sediment surface. However, there may be a DOC-consuming process that is bound by oxidized conditions.

According to balance values given in [51], which were obtained taking into account sediment trap data and oxygen consumption for mineralization of organic carbon at the lake-water-bottom interface, the net burial of particulate organic carbon in sediments is 220 mmol C m⁻² yr⁻¹ in the southern basin and 125 mmol C m⁻² yr⁻¹ in the northern basin. The obtained data on DOC (78 mmol C m⁻² yr⁻¹ in the southern basin and 31 mmol C m⁻² yr⁻¹ in the northern basin) amount to 25–35% of the calculated values shown in [49] and determine the underestimation of burial of organic carbon in bottom sediments. Furthermore, up to 54% of organic carbon has been shown to be sequestered and immobilized in sediments by binding to Fe oxyhydroxides.

Thus, the results of this study provide a deeper insight into the nature and properties of DOC in freshwater ecosystems and compensate for the underestimation of DOC in the internal carbon cycle of the lake.

The bulk of sediments in Lake Baikal are oxidized from the surface, but new areas of oil and gas discharge increasingly appear where sediments are restored from the surface. Such areas also include some littoral zones. The organic carbon cycle processes in these areas will be rather different, so further detailed studies are needed to quantify and investigate the direction of DOC fluxes.

5. Conclusions

The first detailed study of DOC with relatively high resolution in the pore water profile of Lake Baikal revealed that redox conditions are the main regulators of DOC exchange. Oxidized surface sediments that immobilize up to 54% of organic carbon by binding to iron are also an effective trap for DOC where it is sorbed on Fe(III) hydroxides. Further mobilization of DOC in anoxic sediments is correlated with the reductive dissolution of Fe(III) hydroxides. We identified for the first time that DOC fluxes along the profile at the water-bottom interface and in deeper layers of the surface sediments of Lake Baikal change direction twice at the redox interfaces, accounting for 78, −102, and 5.9 mmol C m⁻² yr⁻¹ in the southern basin of Lake Baikal and 31, −33, and 8.6 mmol C m⁻² yr⁻¹ in the northern basin. The detected positive DOC flux (31–78 mmol C m⁻² yr⁻¹) near the water-bottom interface indicates DOC absorption by sediments under oxygenated conditions, i.e., Lake Baikal sediments are a sink of DOC. The obtained data on DOC amounts to 25–35% of the calculated values of the net burial of organic carbon in sediments and determines the underestimation of the burial of organic carbon in bottom sediments. Redox conditions are reversible, and if climate change contributes to these conditions on the lake bottom, it is possible that bottom DOC from sediments will enter the lake water in the future. Further research is needed to quantify the effects of these processes on the DOC flux in sediments.