Research Regarding the Autochthonous Dissolved Organic Carbon to Recalcitrant Dissolved Organic Carbon Transformation Mechanism in a Typical Surface Karst River

Abstract

Autochthonic recalcitrant organic carbon is the most stable component in karst aquatic systems. Still, the processes of its generation and transformation remain unclear, which hinders the study of the mechanisms and quantitative calculations of carbon sinks in karst aquatic systems. This study collected water samples from the Li River, a typical surface karst river in Southwest China. Through in situ microbial cultivation and the chromophoric dissolved organic matter (CDOM) spectrum, changes in organic carbon components and their contents during the transformation of autochthonic dissolved organic carbon (Auto-DOC) to autochthonic dissolved recalcitrant organic carbon (Auto-RDOC) were analyzed to investigate the inert transformation processes of endogenous organic carbon. This study found that microbial carbon pumps (MCPs) promote the tyrosine-like component condensed into microbial-derived fulvic and humic components via heterotrophic bacteria metabolism, forming Auto-RDOC. During the dry season, the high level of Auto-DOC provides abundant organic substrates for heterotrophic bacteria, resulting in significantly higher Auto-RDOC production compared to the rainy season. This study provides fundamental information on the formation mechanisms of Auto-DOC in karst aquatic systems, which contributes to the assessment of carbon sinks in karst aquatic systems.

1. Introduction

In global carbon cycling, previous research estimated that the “missing sink” in terrestrial ecosystems was approximately 1.2 Pg C/a [1]. Karst carbon sink is an important component of the terrestrial “missing sink”, but its role and significance in global carbon budgets have been consistently underestimated [2]; especially, dissolved organic carbon (DOC) is generally neglected in the calculation of the carbon budget based on the assumption that DOC is readily mineralized by planktonic bacteria. Liu et al. [3] demonstrated that a portion of dissolved inorganic carbon (DIC) is converted to autochthonous organic carbon by biological carbon pumps (BCPs), forming a stable carbon sink. The autochthonous organic carbon sink ranges from 0.38 to 1.8 Gt C/a, but uncertainties still exist [4]. The ultimate efficiency of a BCP in forming a stable carbon sink is determined by losses during the stable burial process of autochthonous organic carbon. In karst water systems, microbial carbon pumps (MCPs) predominantly drive the transformation of Auto-DOC to Auto-RDOC, forming a stable carbon sink [5]. However, problems remain with the organic component’s variation and the formation mechanisms of organic carbon components during the Auto-DOC to Auto-RDOC conversion process. Therefore, analyzing the composition and structure of autochthonous organic carbon is crucial for understanding the formation of stable carbon sinks and contributes to a better understanding of carbon cycling in aquatic ecosystems.

DOC accounts for 80% of the total organic carbon content [5], with Auto-DOC representing over 60%. Autochthonic organic carbon is regarded as simple, labile, composed of small molecules, and easily consumed by heterotrophic microbes, which led to a negative budget of autochthonic organic carbon in karst carbon sink [6]. Furthermore, the preferential breakdown and metabolites of autochthonic organic carbon refers to microbial remineralization. In karst aquatic systems, the carbon sink flux of DIC fixation by BCPs to generate Auto-DOC reaches 24.76 tC/(km²·a), but the actual buried flux of endogenous organic carbon is only 37–66% [7]. The mechanism of MCPs dominates the utilization of labile organic carbon within Auto-DOC by heterotrophic bacteria metabolism, resulting in the generation of biologically less available Auto-RDOC, and subsequently leading to the loss of Auto-DOC flux and the increased stability of karst carbon sink [8]. Auto-RDOC exists stably and becomes the recalcitrant carbon sink in karst aquatic systems for its biotic inertness and “dilution effect”, which causes the Auto-RDOC concentration to decrease to below the biologically utilizable threshold (24.63 μmol/L) [9]. However, the components and their contents’ variation remain unclear in the transformation process of Auto-DOC into Auto-RDOC, which is valuable for understanding the MCP mechanism and recalcitrant carbon sink evaluation in karst aquatic systems.

The Li River, which is located in Guilin City, Southwest China, was chosen as an example of a typical surface karst river to investigate the distribution of organic carbon in a karst river system. Samples were systematically collected from the Li River’s main stream and primary tributaries during both the dry and wet seasons. Variation in the intensity of the CDOM’s fluorescence indicated variations in concentrations of different components. The sources and compositional characteristics of DOC during the transformation process of Auto-RDOC in rivers were investigated using in situ microbial cultivation methods. This study aimed to determine the transformation mechanisms of organic carbon components during the degradation of Auto-DOC and the formation of Auto-RDOC in a karst aquatic system. The findings provide valuable insights into the formation and transformation mechanisms of stable autochthonous organic carbon sinks in karst aquatic systems.

2. Study Area

The Li River is a tributary of the Pearl River system located in the central part of Guilin City, Guangxi, China. Its watershed ranges from 24°38′10″–25°53′59″ N and 110°7′39″–110°42′57″ E, representing a typical karst region. The Li River originates from Ma’er Mountain in Yuecheng Ridge, flows southward, and ends at the Sanjiangkou in Pingle, with a total length of 164 km. The Li River watershed exhibits a narrow north–south elongated distribution, with a structural inclination and predominant Devonian and Carboniferous strata at the core of the syncline. The Li River’s main channel is primarily developed near the core of the syncline, which is also the region of concentrated distribution of carbonate rocks.

The Li River watershed showcases a typical karst topography, characterized by thin soil layers with a high content of gravel, discontinuous soil cover, uneven distribution of forests, and segmented features. The upper reaches of the Li River are primarily composed of a lower Paleozoic (Pzl) shallow metamorphic clastic rock system and a red clastic rock series from the lower-to-middle Devonian (D₁–D₂x), which make up the northern low and middle normal mountains with clastic rock characteristics. The middle and lower reaches consist mainly of carbonate rocks from the middle Devonian Donggangling stage (D₂d), upper Devonian Rongxian formation (D₃r), lower Carboniferous Yanhuan stage (C₁y), and Datang stage (C₁d), forming a typical karst landform with peak-cluster depressions (valleys) and fenglin plains in the central and southern regions [10] (Figure 1).

3. Methods

3.1. Sampling and Cultivation

A total of 12 sampling points were defined where river water was collected in April (wet season) and December (dry season) in Guilin (Figure 1). Samples were collected in a 5 L sterile polyethylene bottle after rinsing 3–5 times with water samples. The collected samples were filtered on the same day using a vacuum filtration device (Hilintec-C50, Hailin, China) with a 0.22 μm sterile filter membrane from Millipore. The filtered water was stored in 50 mL brown glass bottles (burned at 300 °C for 3–4 h to remove all organic matter) for CDOM analysis, and kept refrigerated before the experiment.

The in situ water microbiological cultivation method was used for the cultivation of recalcitrant DOC. Fresh water samples were filtered using 0.7 μm glass fiber filter membranes (burned at 300 °C) from Whatman, and then transferred to 1 L sterilized brown glass reagent bottles. Three bottles were filled per sampling point, sealed tightly, and stored at a constant temperature of 20 °C. After 30 days, the RDOC-cultivated water samples were filtered (47 mm mixed membrane, 0.22 μm pore size) and analyzed for CDOM.

3.2. CDOM Spectrum Analysis

Three-dimensional fluorescence spectra were measured using a fluorescence spectrophotometer (RF-5301PC, Shimadzu, Kyoto, Japan) with Milli-Q ultrapure water as the blank. The excitation wavelength range was scanned from 220 to 500 nm with 5 nm intervals, while the emission wavelength range was scanned from 220 to 600 nm with 1 nm intervals. The scanning speed was set to 2400 nm/min.

UV-visible absorbance spectra were obtained using a UV-visible spectrophotometer (UV-2450, Shimadzu, Japan) with ultrapure water as the blank. The scanning wavelength range was from 190 to 800 nm with 0.5 nm intervals.

All tests were conducted at the Key Laboratory of Karst Environment of Southwest University in Chongqing, China.

3.3. Data Statistics and Analysis

Parallel factor analysis (PARAFAC) was employed to analyze three-dimensional fluorescence spectra data using the CDOMFluor toolbox in MATLAB [11]. Analysis results were compared with fluorescence spectra from the OpenFluor open-access database (www.openfluor.org, accessed on 12 August 2023) to determine the types of fluorescent components and their consistency with previous reports.

Fluorescence and ultraviolet-visible absorption spectral parameters can characterize the source, molecular weight, and degree of humification of DOC. Definitions of the fluorescence and ultraviolet absorption coefficients used in this study, as well as their indicative significance, are shown in

Table 1. Definitions of fluorescence and ultraviolet absorption coefficients.

Coefficient	Definition	Meaning
FI	The ratio of fluorescence identities at Ex 370 nm/Em 450 nm and Ex 370 nm/Em 500 nm	FI (Fluorescence Index) values between 1.4 and 1.9 indicate that terrestrial and autochthonous aquatic sources influence DOC sources [12]
HIX	The ratio of fluorescence identities at Ex 254 nm/Em 435–480 nm and Ex 254 nm/Em 300–345 nm	HIX (Humifaction Index) values characterize the content or degree of humification of DOC; a HIX value less than 4 indicates a low degree of humification and poor organic matter stability [13]
SR	The ratio of spectral slopes for two different wavelength ranges (275–295 nm and 350–400 nm), SR = S275–295:S350–400	Represents CDOM sources [14]
a355	The absorbance of CDOM at 355 nm	Represents CDOM contents [15]
SUVA254	The ratio of CDOM absorbance at 254 nm to DOC concentration	Represents the aromatic degree of CDOM [16]
S275–295	The exponential fit of the absorption spectrum from 275–295 nm	Represents the molecular weight of CDOM; a low S275–295 value indicates high molecular weight [17]

.

4. Results

4.1. Organic Matter Characteristic Variations in the Auto-DOC to Auto-RDOC Transformation Process

During the transformation process from Auto-DOC to Auto-RDOC, FI values indicated a shift from a mixed source to a microbial source during the dry season, while samples during the rainy season showed a more pronounced change towards a terrestrial source (Figure 2). The decrease in a355 revealed that microorganisms had consumed some CDOM during the Auto-DOC to Auto-RDOC process. HIX values showed an increase in the degree of organic matter humification and an enhancement of organic matter stability in the water samples. The decrease in S_275–295 indicated an increase in relative molecular weight, consistent with results indicating an increase in the degree of humification, as indicated by the HIX. The slight increase in SUVA₂₅₄ suggested a slight increase in aromaticity.

4.2. Components’ Variation during the Auto-DOC to Auto-RDOC Transformation Process

Five components were identified from PARAFAC, as shown in

Table 2. Characteristics of five fluorescent components analyzed by EEM–PARAFAC in Li River.

Components	Type	Extraction/Emission(nm)		Characteristics
		This Study	Previous Reports
C1	Fulvic-like	255(340)/432	A peak: 240~260/380~460 [18] C2: 335/424 [19]	Substances of lignin degradation metabolism
C2	Fulvic-like	255(290)/383	M peak: 290~310/370~410 [20] C2: < 300/396 [21] C4: 295/358 [22]	Substances derived from microbial metabolism, which was considered as partly degraded semi-labile DOC component
C3	Humic-like	265(370)/480	Posthumic-like: 260~300/475~510 [23] C3: 370(265)/488 [11]	Humic-like substance, the recalcitrant DOC component
C4	Fulvic-like	270(335)/437	C6: 275(330)/436 [24] C1: 270 (355)/446 [25]	Substances of lignin degradation metabolism
C5	Tyrosine-like	275/303	C3: 275/308 [26] C5: 275/302 [27]	Submerged aquatic plant-derived protein-like substance, which is easily degraded labile DOC component

, which included three autochthonous components, C2, C3, and C5, and two allochthonous components, C1, and C4. Among the autochthonous components, C2 was identified as humic substances, categorized as semi-labile DOC; C3 was identified as fulvic acids, categorized as recalcitrant DOC; and C5 was identified as a tyrosine-like substance, categorized as labile DOC.

During the Auto-DOC to Auto-RDOC transformation process, the total fluorescence intensity changed from 58.67 ± 20.92 QSU to 58.54 ± 21.01 QSU in the dry season (p < 0.05) and decreased from 52.70 ± 29.32 QSU to 49.81 ± 29.91 QSU in the wet season (p < 0.01) (Figure 3). There was no significant variation in total fluorescence intensity during the dry season, but a slight decrease in the wet season, which indicated that the organic matter was consumed thoroughly in the wet season.

Three Auto-DOC components were found in the Li River, including C2, C3, and C5. The components C3 and C5 revealed significant increases and decreases in both the dry and wet seasons, but the component C2 increased in the dry season and decreased in the wet season. The component C3 increased by 5.11 QSU in the dry season and 2.33 QSU in the wet season. The component C5 decreased by 13.78 QSU in the dry season and 5.65 QSU in the wet season. The component C2 increased by 3.84 QSU in the dry season and decreased by 1.21 QSU in the wet season.

5. Discussions

MCPs are the key mechanisms that stable carbon sink in terrestrial aquatic systems, which generate and stabilize Auto-RDOC from Auto-DOC. Microbes convert labile Auto-DOC into stable microbial-derived Auto-DOC, which contains complex organic matter structure and resists decomposition. In the Li River aquatic system, the Auto-DOC identified included three main components. These components, sequentially ordered by biological availability, are proteins (C5), humic substances (C2), and fulvic acids (C3) [28]. Protein components consist of simpler molecular structures, like amino peptidases and peptides, with low condensation levels, exhibiting 100% biodegradability, and are consumed by heterotrophic microbes [29]. The protein component C5 was the main labile DOC fraction consumed by heterotrophic microbes, decreasing by 21.44% (dry season) and 10.14% (wet season). Humic substances are the early precursor products formed from protein components during the microbial degradation process, which have smaller molecular weights and simpler structures [30]. Humic acid component C2 can be considered as a semi-labile DOC with certain degradation properties, serving as an intermediate product of the Auto-DOC to Auto-RDOC transformation process. Furthermore, humic substances can polymerize into fulvic acid [31], which is highly recalcitrant with extremely low biological availability (component C3). In the Auto-DOC to Auto-RDOC transformation process, the decrease in C5 and increase in C2 revealed that biological availability reduced and organic carbon became recalcitrant and stable. The proportion of labile DOC decreased by 21.43% and recalcitrant DOC increased by 20.32% in the dry season, while labile DOC decreased by 11.69% and recalcitrant DOC increased by 11.01% in the wet season.

In the dry season, the stable aquatic environment of the Li River promotes aquatic photosynthesis. Consequently, the protein component C5 derived from aquatic plants elevated by 20.27% in the dry season compared to the wet season. Moreover, primary production was inhibited by a water turbidity increase and a radiation decrease induced by soil erosion in the wet season [5], and the allochthonous DOC increased by 10.02%. Heterotrophic bacterial metabolism controls organic matter degradation and recalcitrant transformation in terrestrial aquatic systems [5], which preferentially utilize labile DOC components [32]. Labile DOC revealed higher fluorescence intensity and proportion in the dry season than in the wet season. Heterotrophic microbes primarily degrade and utilize labile DOC in the dry season, so the fluorescence intensity of labile DOC after Auto-DOC to Auto-RDOC transformation decreased to less than 1/3, of which the proportion decreased by about 21.44%. In contrast, because the contents of labile DOC were insufficient in the wet season, component C5 was nearly completely depleted. Thus, strong substrate limitations led to heterotrophic bacteria utilizing semi-labile DOC, which has been described as a “Priming effect” [33]. Due to the “Priming effect”, the relatively stable humic-like component C2 was decomposed too, and the fluorescence intensity of C2 decreased by 1.21 QSU during the wet season (Figure 3).

During the Auto-DOC to Auto-RDOC transformation process, variation in fluorescence components also leads to an overall alteration in the optical properties of CDOM. The relative molecular weight and humification degree increased; the S_275–295 index decreased by 49% and the HIX increased by 89.75%. Because of higher Auto-DOC and labile DOC proportions in the dry season, the humification and aromatization degrees of DOC were lower than in the wet season. However, with the high utilization of labile DOC in the dry season, the variation in humification and aromatization degrees revealed a significantly higher increase in the dry season than in the wet season. The increase in the HIX in the dry season was 2.1 times that of the wet season, and the increase in SUVA₂₅₄ was 1.5 times greater. A significant negative correlation between the C5 component and the HIX indicates (p < 0.05) that the increase in humification primarily resulted from the substantial consumption of low-humified labile DOC [34] (Figure 4). The significant positive correlation between components C1, C3, and the SUVA₂₅₄ index (p < 0.05) suggested that the accumulation of Auto-DOC triggered an increased aromaticity level in the karst aquatic system. The labile DOC was rapidly decomposed in the karst aquatic system, while allochthonous DOC remained a sustained input, which led to a decrease in microbial-derived organic matter [35]. As a result, CDOM spectrum characteristics revealed a variation in which autochthonous indexes FI and SR were negatively correlated with autochthonous components (Figure 4).

6. Conclusions

In the Li River aquatic system, the Auto-DOC included three CDOM components: the protein-like component C5, microbially derived humic-like component C2, and fulvic acid-like component C3. Changes in the components of CDOM after recalcitrant DOC cultivation demonstrated that MCPs primarily operate with protein-like components as a foundational substrate, humic-like components as intermediate products of Auto-DOC, and fulvic acid-like components as ultimate products of recalcitrant DOC. High primary productivity was found in the dry season, which generated a high level of Auto-DOC and offered a sufficient nutrient substrate for heterotrophic bacteria. The high content of labile DOC simultaneously reduced the “Priming effect” and was beneficial for the accumulation of recalcitrant DOC.

This study revealed the transformation mechanisms of Auto-DOC to Auto-RDOC in karst surface rivers and the main factors that influence the Auto-RDOC formation process, providing theoretical support to the critical role of Auto-DOC in karst carbon sink and offering foundational information for carbon sinks evaluation in karst surface aquatic systems.