Coupled nitrogen transformation and carbon sink in the karst aquatic system: a review

The term ‘karst’ refers to regions that typically developed from carbonate rocks. These rocks mainly include mineral calcite (CaCO₃) and mineral dolomite (CaMg(CO₃)₂), containing limestone and dolomite rock or dolostone, respectively (Hartmann 2015). Carbonate bedrock regions represent 14% of Earth's continental surface, including broad swaths of southwestern China. The southwestern China karst region is one of the largest globally continuous karst areas, covering ∼540 × 10³ km² over eight provinces (Zhang et al. 2020). Carbonate karst aquifers serve as the drinking water source for about one-quarter of the global population (Ford & Williams 1989).

Stress on climate change due to carbon dioxide (CO₂) emissions has increased significantly in recent decades, which is attributed to deforestation, the change of land-use type, and the use of fossil fuels (Melnikov & O'Neill 2006). Liu et al. (2015) estimated that China's cumulative carbon (C) emission during 2000–2013 was nearly 2.9 gigatons, which was larger than China's estimated total forest sink during 1990–2007 (2.66 gigatons of C). Moreover, this estimated C sink was still undervalued in terms of deserts and karst formations. In China, two considerable challenges for estimating C sink are the amount of CO₂ emitted and absorbed by the landscape. As limestone degradation could be a substantial inorganic C sink, the C sink in the karst area might play an important role in the global C cycle and the balance of China's C emission.

Water plays a critical role in the C cycle in the karst area, as it is a basic medium to transform derived organic and inorganic carbon, and C can also be exchanged with the atmosphere in the form of CO₂ across the water–air interface. Previous studies have shown that many non-karst (mainly silicate area) rivers are usually supersaturated with CO₂, then leading to the emissions of CO₂ (Butman & Raymond 2011; Wang et al. 2011; Rasilo et al. 2017). However, in karst water, such as rivers, lakes, and reservoirs, carbonate mineral weathering consumes CO₂ with a more rapid ratio and, therefore, increases the concentrations of dissolved inorganic carbon (DIC) (mainly ⁠) (Berner 2003; Liu et al. 2011). As DIC can be consumed by aquatic phototrophs by photosynthesis and transformed to organic C in inland water, the absorption of CO₂ continues and forms karst C sink rapidly (Liu et al. 2011). Therefore, it is of great importance to understand the role of karst processes as a global C sink and the quantification of C fluxes.

Intensive agriculture and urbanization have led to the excessive input of N into the soil and further increased the amount of nitrate (NO₃⁻) in water (Xin et al. 2019), as well as changed the N cycle process in the aquatic ecosystem. Human activities have delivered 64 Tg N/yr to rivers and streams in the 20th century, which doubled that in the 19th century (34 Tg N/yr) (Beusen et al. 2016). It has been shown that the cycling of C and N in karst aquatic systems is closely related to each other, as coupled C–N cycling is involved in the transformation of DIC into dissolved organic carbon (DOC) (Zhao et al. 2021). The increasing anthropogenic atmospheric CO₂ emissions result in a progressive increase in N limitation in plants. C sequestration and increased atmospheric CO₂ concentrations would lower terrestrial N availability and lead to decreases in N flux to the atmosphere and N deposition to aquatic ecosystems. As industrial and agricultural discharge drives a sharp increase in anthropogenic N emissions, the magnitude of N from this anthropogenic input will likely become large enough to sustain similar conditions of ecosystem N availability (McLauchlan et al. 2013). The elevated DIC concentrations in karst water may also enhance the aquatic photosynthetic uptake of DIC (Liu et al. 2010b, 2011) and then may promote N translation. However, the interactions between N translation and CO₂ absorption in karst aquatic ecosystems have not been well depicted.

In this review, we provide an overview of the relevance of karst regions to the special water characteristics and discuss the impact on the N transformation due to these specific water characteristics. We (1) start with an introduction to the characteristics of karst water, (2) present an overview of N transformation in karst water systems, (3) discuss the processes related to carbon sink, and (4) show the challenges and new directions in karst water C/N balances.

The carbonate weathering is formulated as CaCO₃ + CO₂ + H₂O ↔ + Ca²⁺ and CaMg(CO₃)₂ + 2CO₂ + 2H₂O ↔ 4 + Ca²⁺ + Mg²⁺. CO₂ consumed during carbonates dissolution would finally be released to the atmosphere by the precipitation of carbonates in the oceans (Berner et al. 1983), and the kinetics of dissolution is much faster as compared to precipitation considering the ocean turnover time (timescales of <3 ka). Therefore, carbonate weathering is critical in global CO₂ balances (Van Cappellen & Qiu 1997; Kump et al. 2000).

CO₂ was absorbed by raindrops formed in the atmosphere, and of which the concentration would further increase after precipitation and infiltration in the soil due to vegetation and microbial processes. Furthermore, the CO₂ in soil water could dissolve the bedrock underlying soil, composed of carbonate rock, during percolation of soil moisture in the karst area. Previous research found that the contribution of carbonates, e.g. ⁠, to the total dissolved load in the lakes and rivers worldwide was up to 38% (Ferris et al. 1994).

The rapid process of carbonate weathering resulted in remarkably higher concentration of bicarbonate and calcium in water in carbonate terrains as compared to silicate terrains (Liu & Dreybrod 1997; Liu et al. 2007; Raymond et al. 2008). Therefore, the reaction between carbonate minerals and CO₂, which increases DIC (DIC = CO₂(aq) + + ⁠) concentrations, may impact the C cycle and represent a net sink of atmospheric CO₂ in a short time scale (Martin 2017).

Previous studies in Table 1 showed that the karst water pH values ranged from 6.45 to 9.7, with pH in most of the water samples higher than 7. Considering dissolved CO₂ is mainly present as (Liu et al. 2018) when the water pH varies from 6.5 to 10, the main form of DIC in karst water is and the conversion of to CO₂ is slow. While the concentration varies in different karst water samples, with the values ranging from 12.2 to 2,633 mg/L, the concentration in karst water is much higher than that in non-karst water of similar environmental parameters (e.g. temperature and precipitation). In addition, great variations of Ca²⁺ and Mg²⁺ concentrations were also observed. In terms of geographic locations, the concentrations of ⁠, Ca²⁺, and Mg²⁺ in South China were higher than those in North China, which could be attributed to the differences in environmental parameters, such as temperature, lithology, and climate. As can be seen in Table 1, the concentrations of ⁠, Ca²⁺, and Mg²⁺ in a typical karst catchment in Guangxi, China were about 10 times higher than that in the Yellow River. Therefore, it is summarized that karst water is typically with high contents of ⁠, Ca²⁺ and Mg²⁺, and high pH.

As shown in Table 2, the organic carbon (OC)/N ratio in karst water in the upstream of the Pearl River (11.8:1) is about twice compared to that in the Pearl River estuary (5.0:1) (Liu et al. 2020). While the organic carbon can be categorized into DOC and particulate organic carbon based on a size threshold of 0.2 μm. A previous study showed that the DOC/N ratio in a karst reservoir in Guilin, China was 1.11:1 (Song et al. 2017), which is not consistent with aforementioned Liu et al.'s work. Moreover, the OC/N ratio values in karst soil were slightly lower than those in non-karst soil (Gu et al. 2018). In addition, Table 2 also shows that the DOC/N ratios in rivers of China are much lower than those in the Mississippi River estuary and Yenisei River, probably due to high precipitation and high organic matter (OM) inputs, indicating that the OM might be the key factor influencing the DOC/N ratio in rivers as riverine OM is with high DOC/N ratio (about 30:1) (Bauer et al. 2013).

Neither the data of the DIC/N ratio in the karst or the non-karst aquatic systems are available. The DIC contents in karst water are relatively higher as discussed before; therefore, it can be deduced that the total concentrations of carbon including DIC and DOC in karst water are also higher than those in non-karst water. As organisms translate DIC into total organic carbon (TOC) by photosynthesis, theoretically the total carbon (TC)/N ratios in karst water should also be higher. Considering the limited number of researches available currently, more researches are still required to figure out the differences of TC/N ratios between karst and non-karst water.

As one of the most important cycles in water systems, the N cycle, including the conversion and flux of N, has received worldwide attention in recent years. Human activities, such as intensive agriculture and rapid urbanization, lead to excessive and repetitive N inputs, which significantly affect the natural cycle of N. In the 20th century alone, human activities have increased the amount of N delivering to rivers and streams from 34 to 64 Tg N/yr (Beusen et al. 2016). N pollution has become one of the most concerned and prevalent environmental problems, especially in karst areas. Karst areas are subjected to greater pressure of N pollution than other regions because karst aquifers are particularly sensitive and vulnerable to chemical pollution from human activities due to their developed networks of fractures, pipelines, and sinkholes (Jiang 2013).

In water systems, the two sources for N are natural and artificial activities. While biological N fixation was the only important process in Earth's ecosystems producing reactive N to support C fixation into energy-rich OM (primary production) before the agricultural and industrial revolutions, human activities have been introducing large amounts of N into the environment through municipal sewage, industrial effluent, and agriculture since the agricultural and industrial revolutions (Wakida & Lerner 2005), especially the N fertilizer used in agriculture to promote crop growth and increase crop yield (Wang et al. 2019a, 2019b).

In aquatic environments, N was composed of organic and inorganic N (Nie et al. 2018). The organic nitrogen is divided into dissolved organic nitrogen (DON) and particulate organic nitrogen (PON), depending on whether it can pass through a 0.2-μm filter (Jørgensen 2009). DON includes a variety of organic molecules and compounds, ranging from small molecules like urea and amino acids, to peptides and proteins, while PON includes both dead OM and living organisms that are larger than 0.2 μm (Jørgensen 2009). The inorganic nitrogen in aquatic ecosystems includes dissolved N₂ gas, oxidized ions such as nitrate (NO₃⁻), nitrite (NO₂⁻), ammonium ion (NH₄⁺), and ammonia gas (NH₃) (Howarth 2009). The most frequently detected mineral N fractions in water are NO₃⁻ and NH₄⁺, which are also the dominant components of N produced by human activities (Beusen et al. 2016).

In water systems, N would go through a variety of bacteria-mediated processes, mainly including nitrogen fixation, mineralization, nitrification, denitrification, dissimilated nitric acid reduction to ammonium, and ammonia oxidation (Figure 1). Most N on Earth is in the form of N₂, which becomes biologically significant after being fixed by bacteria, lightning, volcanic activity, and human activity. In aquatic environments, N fixation is mostly carried out by heterotrophic or autotrophic bacteria and cyanobacteria. NO₃⁻, NO₂⁻, NH₄⁺, and NH₃ are the active N in water. Algae, rooting plants, fungi, and bacteria absorb and reduce NO₃⁻ and NO₂⁻ to NH₄⁺ in a process known as assimilative nitrate or nitrite reduction. NH₄⁺ could be catalytic-oxidized by nitrifying bacteria to NO₃⁻ in a process called nitrification, from which the nitrifying bacteria gain energy to fix CO₂ into new bacterial biomass. Plants, algae, and microorganisms use nitrates and ammonium to produce organic nitrogen-containing compounds, which could be further taken up by animals through the food chain, through one of the following processes: direct absorption, assimilation, and reduction. The organic N eaten by animals or decomposed by microorganisms is excreted as ammonium or as urea which is further rapidly hydrolyzed to ammonium. In addition, NO₃⁻ is also reduced by heterotrophic or autotrophic bacteria to NO₂⁻, which is further reduced to N₂ by a process known as traditional denitrification or dissimilated nitrate reduction. These processes that release N from organic N back to the environment are called nitrogen mineralization. The other N cycle processes include denitrification based on chemosynthetic oxidation of sulfides or reduced iron (Howarth 2009), anaerobic oxidation of ammonia to N₂ (Anammox), dissimilatory reduction of NO₃⁻ to NH₄⁺ via NO₂⁻ (DNRA) (Medinets et al. 2015), and autohydrogenotrophic denitrification of NO₃⁻/NO₂⁻ to N₂. However, the relative importance of these newly discovered processes in water systems remains quite uncertain (Howarth 2009).

Due to the unique hydrochemistry of karst water (such as high Ca²⁺ content, high pH, and high DIC concentrations), the N cycle characteristics in karst areas are different from those in non-karst areas. In the perspective of the N cycle, the alkalinity and high DIC contents in karst areas can inhibit the heterotrophic denitrification and mineralization processes, which determine the karst aquatic environment as an N sink. During nitrification, the oxidation of every ammonium ion produces two protons worth of acidity and makes the environment more acidic (Howarth 2009). Therefore, the alkaline environment in karst water is beneficial to neutralizing the acid generated by nitrification and further promotes nitrification. However, the denitrification process is opposite to nitrification. Every nitrate ion consumed during denitrification consumes one proton of acidity, and thus, this process tends to raise the pH of the environment (Howarth 2009). Consequently, the inherent alkalinity of karst water will inhibit the denitrification reaction. Furthermore, the high DIC/N ratios in karst water can inhibit the gross mineralization because microbes immobilize rather than mineralize N to maintain the stoichiometric ratio of DIC/N in their biomass (Xin et al. 2019).

The cycling of C and N in aquatic environments is closely related, with coupled control of organic carbon concentrations through aquatic biological processes of assimilation or denitrification (Gruber & Galloway 2008; Zeng et al. 2019). Although there are few studies on the C–N coupling cycle in karst water systems, some studies have found that excessive nitrogen emissions from human activities lead to the C–N coupling cycle participating in the carbonate weathering process, resulting in the increase of DIC flux in karst water systems (Raymond et al. 2008; Jiang 2013; Zhao et al. 2020). Most biological processes in water systems are C-limited processes, but this is not the case in karst aquatic environments. As shown in Figure 2, in karst water systems, higher DIC and increased NO₃⁻ concentration due to the enhanced nitrification and human activities can promote the growth of aquatic communities (Liu et al. 2018; Zeng et al. 2019). The growing amount of algae and microorganisms in water increases the consumption of DIC and NO₃⁻ through photosynthesis, as the conversion of DIC to OC by photosynthesis induces the consumption of NO₃⁻, and therefore, reduces the NO₃⁻ concentration (Pedersen et al. 2013; Nõges et al. 2016; Liu et al. 2018). In addition, during this process, DIC and NO₃⁻ are converted to OM and O₂ is released, which contrasts with the traditional knowledge that CO₂ is released during carbonate precipitation (Jiang 2013). It has been shown that in the Lijiang River water, the consumption of DIC and NO₃⁻ by aquatic photosynthesis was in a ratio of 9:1 (mol/mol) to produce autochthonous DOC (Zhao et al. 2021). To sum up, the C–N cycle coupled with DIC and NO₃⁻ promotes the generation of in-situ DOC in karst aquatic environments, which constitutes the relatively long-term natural C and N sinks in karst water systems, as shown in Figure 2.

Five major components in the global C budget are fossil CO₂ emissions (E_FOS), emissions from land-use change (E_LUC) (mainly deforestation), atmospheric CO₂ (G_ATM), ocean CO₂ sink (S_OCEAN), and terrestrial CO₂ sink (S_LAND). Over the last decade (2010–2019), E_FOS and E_LUC were 9.6 ± 0.5 and 1.6 ± 0.7 Pg·C/yr, respectively, G_ATM was 5.1 ± 0.02 Pg·C/yr, S_OCEAN and S_LAND were 2.5 ± 0.6 and 3.4 ± 0.9 Pg·C/yr, respectively, and the imbalance budget (B_IM) was −0.1 Pg·C/yr (Friedlingstein et al. 2020). The accepted values for S_LAND at present range from 1.8 to 3.4 Pg·C/yr in the global C budget (Melnikov & O'Neill 2006; Lal 2008; Friedlingstein et al. 2020). Similarly, another research indicated that S_LAND was calculated at 2.6 Pg·C/yr, while CO₂ sink due to photosynthesis and CO₂ emissions from plant respiration were 14.1 and 11.6 Pg·C/yr, respectively (Yuan & Liu 2003).

A large C sink is missing from the global carbon cycle with the value of 1.7–2.5 Pg·C/yr (Cao et al. 2011). CO₂ sink from carbonate might be an important component missed in previous studies; however, the volume of CO₂ sink by carbonate weathering on continents varied greatly in different researches, with its volume ranging from 0.018 to 0.6 Pg·C/yr worldwide, which is about 7–36% of the missing C sink (Yuan 1997; Gaillardet et al. 1999; Gombert 2002; Liu et al. 2010b). Liu et al. (2011) found that carbonate weathering contributed about 94% to the atmospheric CO₂ sink, while only 6% resulted from silicate weathering. More researches on CO₂ absorption experiments using more accurate calculation methods are still needed to estimate the atmospheric CO₂ sink by carbonate weathering.

Two primary processes related to global sinks of atmospheric CO₂ are the transformation of CO₂ to in water due to rock weathering (Li et al. 2011) and the assimilation of CO₂ by photosynthesis to organic C (Sabine et al. 2004; Zeebe & Caldeira 2008). There are three sources of ⁠, including carbonate weathering by carbonic acid, carbonic weathering by sulfuric and/or nitric acids, and silicate weathering. Carbonates can consume more atmospheric and petrologic CO₂ because of their high solubility and high dissolution rate (Liu et al. 2010b), resulting in aquatic photosynthesis is the main process for CO₂ absorption in silicate terrains. The roles of both carbonate mineral weathering and aquatic photosynthesis, termed as coupled carbonate weathering, are significant in karst areas (Liu et al. 2018). In addition, Sun et al. (2021) also observed that the average contributions made by silicate weathering to the CO₂ sink in the Lijiang River basin ranged from only 2.3 to 14.8%, which indicated that carbonate weathering was the main source of in this basin although carbonate rock area (3,832 km²) is smaller than silicate rock area (5,482 km²).

DIC is mainly consumed by phototrophs in aquatic ecosystems such as rivers, lakes, and the oceans (Dean & Gorham 1998; Cassar et al. 2004; Tortell et al. 2008). During this process, DIC is transformed in the water to OC and pCO₂ is reduced, which results in the continuous uptake of atmospheric CO₂ (Liu et al. 2010b). The biological productivity of aquatic phototrophs has been found to be associated with the supply of DIC from rock weathering. For instance, the utilization of by Oocystis solitaria Wittr in karst water was 4.6-fold higher than that in non-karst water (Liu et al. 2010a).

Previous researches concluded that the atmospheric CO₂ consumed by carbonate weathering was compensated by a CO₂ released from marine carbonate precipitation over a relatively short time. However, this conclusion neglected the large amount of atmospheric CO₂ uptake during aquatic photosynthesis (Liu et al. 2010b). Moreover, Liu et al. (2018) also observed that the increased DIC concentration controlled by carbonate weathering in the karst area might enhance aquatic photosynthesis, promoting the consumption of atmospheric carbon. Therefore, the amount of CO₂ absorbed by the C sink in the karst area was larger as compared to other areas.

As shown in Table 3, the range of carbon sink rate in carbonate area was 1.54–73 tC/km²/yr, while in silicate area the range of this value was 0.02–8.0 tC/km²/yr. The annual average C sequestration flux of limestone weathering in China was estimated to be 4.28–5.02 tC/km²/yr (Li et al. 2019). The C sinks produced by carbonate weathering and the ‘biological C pump’ in the Li River basin were 12.17 and 2.24 tC/km²/yr, respectively (Sun et al. 2021). The average amount of CO₂ consumed by C sinks in karst basins around the world (8.5 tC/km²/yr) was about three times higher than that in non-karst basins (2.86 tC/km²/yr). In karst water, the C sink is enhanced by N translation, as the concentration of dissolved CO₂ was decreased dramatically (about 75%) due to the increase of NO₃⁻ in the Yangtze River (Wang et al. 2007). Therefore, in the karst basin, the transformation of C and N by aquatic phototrophs was coupled.

Carbonate bedrock regions represent 14% of the continental surface of the Earth and provide drinking water resources for about 25% of the global population. C sink in karst water plays an important role in the global C cycle due to the CO₂ consumption during carbonate mineral weathering. This review highlights the consumption of CO₂ during carbonate weathering and the relevance of the N and C cycles in karst regions. Existing concentrations of ⁠, Ca²⁺, Mg²⁺, and C/N ratios are presented, followed by a review of previous studies.

Here we show that,

1. Karst aquatic systems are characterized by high contents of ⁠, Ca²⁺, Mg²⁺, high pH, and high C/N ratios.

2. The cycling of C and N in karst aquatic environments is closely related and the high content of DIC in karst systems will increase the nitrogen sink. Also, the growth of aquatic communities was promoted by higher concentration of DIC and the increase of NO₃⁻ due to the enhanced nitrification and human activities.

3. The budget of CO₂ sink from carbonate ranged from 0.018 to 0.6 Pg·C/yr (about 7–36% of the missing carbon sink), which indicated that carbonate weathering was an important component neglected in previous studies.

4. The range of C sink rates in carbonate area was 1.54–73.00 tC/km²/yr, while in silicate area the range of this value was 0.02–8.00 tC/km²/yr.

The N input in karst areas worldwide would increase due to continuous population growth, industrial activities expansion, and lifestyle improvement. This necessitates the collection of sufficient information about the C and N sinks in karst systems to provide reliable projections of C and N balances. As few studies are available specifically focusing on the impact of karst characteristics on the C sink as well as the N cycle, more field observations targeting the relationships between C and N cycles in karst aquatic systems are urgently needed to figure out the mechanism of N sink in karst water with high abundance of DIC.

This work was supported by the National Natural Science Foundation of China (No. 42067048, 41807387, and 42007310), Guangxi Foundation for Program of Science and Technology Research (No. 2020GXNSFAA159169), and Guangxi ‘Bagui Scholar’ Construction Project and Guangxi Science and Technology Planning Project (No. GuiKe-AD18126018).

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