The Jiaozhou Bay (120º03°-120º25° E, 36º01°-36º15° N) is a semi-closed bay located in the central part of the Yellow Sea and on the south coast of the Jiaodong peninsula. The existing coastal wetlands along the coast are mainly distributed in the north and northwest of the peninsula, with a total area of 348 km². It is the largest wetland along the estuarine bay in the Shandong Peninsula and has been included in the “national important wetland directory”. The coastal estuary wetland of Jiaozhou Bay and its vicinities are in the warm temperate continental monsoon climate zone, which is regulated by the ocean monsoon. The annual average temperature is 12.2°C, and the annual average rainfall is approximately 900 mm. The annual frost-free period is about 220 d. The tide is a typical semidiurnal tide with an average tidal range of 2.71 m and a maximum tidal range of 6.87 m. The flood tide duration is less than that of the ebb tide. Jiaozhou Bay is the mother bay of Qingdao, linking five main rivers including Dagu River, Jiaolai River, Yanghe River, Baisha River, and Ink River. The experimental area is the estuarine tidal flat wetland located in the middle and lower reaches of the Dagu River and the Yanghe River in Jiaozhou Bay. Among them, the Dagu estuarine tidal flat wetland is located in the middle and lower reaches of the Dagu River Basin, where the wide tidal flats, with fertile water quality and suitable salinity, have been developed into a typical aquaculture area (mainly including prawns and clams). The Yanghe River is a large river flowing into the Jiaozhou Bay, with a total length of 49 km and a basin area of 303 km² (Xi et al., 2017). Since the introduction of Spartina alterniflora from abroad in 1979, a typical Spartina alterniflora marsh has formed in the Yanghe River estuary, and serves as a representative area for the study of alien species invasion.

According to the hydrogeological conditions and the distribution of vegetation in the Jiaozhou Bay, the field soil samples were collected in the study area in July 2016, and four representative wetland types, including mudflat, river flat, Spartina alterniflora marsh and aquaculture ponds (Fig.1), were selected to analyze the characteristics of SIC reserves distribution. Each wetland type was set up with a different number of sampling sites. Specifically, the first group of samples was collected from the mudflat along the coastal zone (No. 1-7); the second group was collected from the river flat along the vertical coastal zone (No. 8-14, with distances of 0, 0.3, 0.7, 1.1, 2, 2.8 and 3.5 km from the sea, respectively); the third group was collected at the Dagu estuary and the Yangkou estuary, including aquaculture ponds (No. 15-17), nearshore Spartina alterniflora marsh (No. 18), and the surrounding mudflat (No. 19), the far shore Spartina alterniflora marsh (No. 7, No. 7 is part of the first group of samples, but is also part of the Spartina alterniflora marsh), and the surrounding mudflat (No. 20). A total of 20 sample sites were selected in the three parts. Three parallel soil profiles at depths of 0-20 cm, 20-40 cm, and 40-60 cm at each sample site were collected. After air-drying, the parallel soil profiles were mixed with the same depth to represent the corresponding soil sample. All soil samples were placed in zip-lock bags immediately and transported to the laboratory. After drying at room temperature, the visible plant and animal residues were removed from the soil samples. Additionally, all soil samples were sieved, and then stored prior to the next processing step.

SIC includes HCO₃^- in soil solution, CO₂ in the soil air, and CaCO₃ deposited in soil, but CaCO₃ is the dominant one present in large quantities. The content of CaCO₃, salt content, and soil bulk density were determined according to the Methods of Soil Agricultural Chemical Analysis (Lu, 2000). The CaCO₃ in the soil was measured through the gas volume method. The salt content was determined through the mass method. The soil bulk density was measured through the cutting ring method. All the samples were prepared as 1:5 soil:water ratio extracts.

SIC content is calculated by the following formula:

$\text{ }\!\!\omega\!\!\text{ }CaC{{O}_{3}}=\frac{\left( V-{{V}_{0}} \right)\times r\times 2.27}{m\times {{10} ^{6}}}\times 100\% (1)$

where V is the volume of carbon dioxide, mL; r is the density of carbon dioxide, g mL^-1; 2.27 is the coefficient of mass conversion of carbon dioxide to calcium carbonate; and m is the weight of the soil sample, g.

The SIC reserves are calculated through the method of stratified cumulative summation calculation as follows:

$SIC=\sum\limits_{i=1}^{n}\, SI{{C}_{i}}\times {{D}_{i}}\times {{H}_{i}} (2)$

where SIC is the soil inorganic carbon reserves, kg m^-2; i represents the soil of layer i; n is the number of soil layers; SIC_i is the soil inorganic carbon content of layer i, g kg^-1; D_i is the bulk density of layer i, g cm^-3; and H_i is the thickness of layer i, cm.

The SIC reserves of the four representative wetland types are calculated based on formula (2). Descriptive statistics of SIC reserves in different types of wetlands are shown in Table 1.

Microsoft Excel 2010 was used to collate the experimental data, SPSS 17.0-Duncan analysis was used to test the difference of data between groups, and Origin 8 and CorelDRAW 12 were used to draw figures.

3.1.1 Mudflat

The variation coefficient (CV) of SIC reserves in the mudflat is 13.97% (Table 1). The soil variability is divided according to the CV (Zi et al., 2016), and the variability of the mudflat is characterized as moderate variability. Therefore, the spatial distribution of SIC reserves in the mudflat is substantially different.

In the horizontal direction (Fig. 2), the differences of SIC reserves of the mudflat are significantly different. The SIC reserves of No. 1-5 sample sites in the Dagu river tidal flat wetland and No. 6-7 sample sites in the Yanghe River tidal flat wetland are all at the lowest values in the estuary, suggesting that the farther away from the estuary a site is, the larger its SIC reserves are. As for the distribution trend, the SIC reserves of Dagu River tidal flat wetland are larger than those of Yanghe River tidal flat wetland on the whole.

3.1.2 River flat

The variation coefficient of SIC reserves in the river flat is 22.08% (Table 1). According to the result of the division, the variability of the river flat is characterized as moderate variability. This suggests that the spatial distribution of SIC reserves in the river flat is quite different.

In the horizontal direction (Fig. 3), the SIC reserves of the river flat at different distances from the sea are not significantly different. The soil stored at the sea entrance (0 km from the sea) has the highest inorganic carbon. As the distance from the sea increases, the reserves of SIC decrease, and they achieve the minimum value at 0.7 km from the sea. Then at greater distances, the SIC reserves increase with the increase in distance from the sea and reach the second- highest value at the farthest distance from the sea (3.5 km).

3.2.1 Mudflat

In the vertical section (Fig. 4), the difference of SIC reserves among different soil layers is not obvious, indicating that the SIC reserves are not closely related to the soil depth.

With the deepening of the soil layer, the SIC reserves of the mudflat first decrease and then increase. The SIC reserves in 0-20 cm, 20-40 cm and 40-60 cm depths account for 33.21%, 32.68% and 34.11% of the total SIC reserves in 0-60 cm, respectively, suggesting that the deep soil is the main place for SIC sequestration in the mudflat.

3.2.2 River flat

As was found for the mudflat, there is no significant difference in SIC reserves among different soil layers in the river flat (Fig. 4). The SIC reserves in 0-20 cm, 20-40 cm and 40-60 cm depths account for 38.18%, 29.73% and 32.09% of the total SIC reserves in 0-60 cm, respectively. Here, the SIC reserves in the surface soil are the highest.

3.3.1 Spartina alterniflora marsh

The variation coefficient of SIC reserves in Spartina alterniflora marsh is 17.41%. According to the result of the division, it falls in the moderate variability category, representing the heterogeneous spatial distribution of SIC reserves.

In the horizontal direction (Fig. 5), the effect of species invasion on the SIC reserves of the wetland is analyzed by comparing SIC reserves between Spartina alterniflora marsh and the surrounding mudflat in the Yanghe River estuary. The SIC reserves of the Spartina alterniflora marsh and surrounding mudflat can be ranked as follows: the far shore mudflat > the far shore Spartina alterniflora marsh, and the near shore mudflat > the near shore Spartina alterniflora marsh. The overall distribution trend shows that the SIC reserves in the mudflat are larger than that in the Spartina alterniflora marsh, and the SIC reserves of both of them are significantly affected by the distance from the sea.

In the vertical section (Fig. 6), the SIC reserves in 0-20 cm, 20-40 cm and 40-60 cm depths account for 33.81%, 29.53% and 36.66% of the total SIC reserves in 0-60 cm, respectively. The deep soil is the main area for SIC sequestration.

3.3.2 Aquaculture ponds

The variation coefficient of SIC reserves of aquaculture ponds is 8.59%, which falls in the range of weak variation. The SIC reserves in the aquaculture ponds have a small spatial distribution difference.

In the horizontal direction (Fig. 7), the SIC reserves of the aquaculture ponds and surrounding mudflat of Dagu River estuary are compared to analyze the effects of anthropogenic activities on the reserves of SIC. The results show that the SIC reserves in the aquaculture ponds are higher than the SIC reserves of the surrounding mudflat on the whole. In the vertical section (Fig. 6), the SIC reserves in 0-20 cm, 20-40 cm and 40-60 cm depths account for 34.81%, 29.87% and 35.31% of the total SIC reserves in 0-60 cm, respectively. More SIC is stored in surface soil and deep soil than in intermediate depth soil.

The mudflat and the river flat are both located in the delta region of the river flowing into the sea. It is a typical estuarine wetland which is formed by the interaction of sea and land, and it is strongly affected by the tidal effects of the ocean. The hydrologic process is thus an important factor for SIC reserves. As the media between inshore and coastal wetland ecosystems, tidal effects can contribute to transverse material migration and energy transformation between estuarine, wetland and coastal areas (Bauer et al., 2013). Under the effects of hydrological processes, the large water flow at the entrance to the sea and the strong scouring effects facilitate the reaction between CO₂ in soil and seawater. As a result, the generated H₂CO₃ decreased the pH, causing the dissolution of CaCO₃ and reduction of SIC (Yang et al., 2010). Moreover, suspended sediment will carry a significant amount of material during the tidal period, which will be deposited under flocculation. At distances farther away from the estuary, the flow velocity becomes smaller. Nevertheless, the strength of the water disturbance cannot resuspend the sediment in the flat surface, causing the accumulation of SIC, and this effect is the main factor leading to the variation of SIC reserves in the horizontal direction of the mudflat (Duan et al., 2008). In the vertical section, there is no vegetation cover in the light beach, and the interference intensity of the ocean tides is greater. This not only affects the lateral migration of the land and water, but also accelerates the deposition rate of the material. The Ca²⁺ of the soil surface is combined with the carbonated phase produced from the dissolved CO₂ in the air or the decomposition of the plants and plant residues to form Ca(HCO₃)₂, which then moves downward by seawater leaching, and is deposited in the middle and lower soil profiles in the form of calcium deposits (Mi et al., 2008).

Salinity is an important ecological factor in the estuarine environment. It will directly or indirectly affect many physical, chemical, biological and biogeochemical processes. Frequent periodic seawater tides will cause changes in soil salinity (Liu et al., 2012). The carbon sequestration is achieved by affecting plant growth and vegetation growth structure in wetland ecosystems. Yu et al. (2015) confirmed that salinity is an important hydrochemical factor affecting inorganic carbon in surface sediments. Therefore, we measured the soil salt content of the 7 sampling sites of the river flat, and analyzed the relationship between the SIC reserves and the soil salt content. The results showed that both SIC reserves and soil salt content present similar distribution trends with distance from the sea on the whole. However, SIC reserves first increased and then decreased with the increase of soil salt content. This phenomenon indicates that the salt content below a certain critical value, the increase of salt content will provide the required ions for the formation of SIC, and above that value, it will affect the processes of respiration, digestion and decomposition of plants and microorganisms in the soil, thus reducing the activity of soil organisms and slowing down the process of organic matter to inorganic matter transformation in the soil, and inhibiting the production of carbonate (Setia et al., 2011; Li et al., 2017; Zhao et al., 2017). These trends are not only related to the dissolution of seawater in estuaries, but are also related to the mineralization phenomena caused by anthropogenic activities. Because of the factories, residential areas and other buildings located in the inland part of the region, the disturbances to wetland soil are obvious, leading to the increase of exogenous substances entering the surface soil, the acceleration of the deposition rate, and the deposition of SIC (Duan et al., 2008). In addition, studies have shown that inorganic carbon content in sediments is affected by different material sources, sedimentary environments, redox processes, and other factors. Moreover, the physical and chemical coagulation, such as the marine redox conditions, the organism metabolism and the physical and chemical agglomerates, also affect SIC reserves in offshore areas (Yates et al., 2006). Therefore, the changes of SIC reserves in the river flat reflect the effects of both hydrologic and anthropogenic factors on SIC migration, and have great complexity.

Spartina alterniflora is the most representative invasive species in Jiaozhou Bay. The mulching effect of plant leaves and the water conservation effect of roots can effectively reduce water evaporation, increase soil moisture content, and form an environment of soil anoxia (Tang et al., 2014). This anoxic environment can cause roots to be exposed to the surface, which weakens the flow velocity of the tidal water and deposits the carried SIC. Then plants transform the SIC into organic matter through photosynthesis, decreasing the reserves of SIC. The growth of vegetation is also an important factor affecting the carbon sequestration capacity (Bernal et al., 2012). Studies have found that short-term Spartina alterniflora invasion will decompose carbon pools for plant uptake and utilization (Yang et al., 2010), which results in low SIC reserves in the soil of nearshore Spartina alterniflora marshes. Through these processes, differences in plant growth characteristics and years can lead to variations in SIC reserves.

The planting of Spartina alterniflora can change soil structure and soil colloid state (Liu et al., 2011). The wetland soil transition layer is generally about 30 cm underground; and it is hard for the plant roots to penetrate into deeper soil where fewer roots can be found (Liu et al., 2003). Accordingly, the densely packed soil layer has low SIC reserves. In addition, the decomposition of SOC by Spartina alterniflora promoted the dissolution of inorganic carbon in surface soil, which caused SIC to move down with the increase of soil depth (Pan et al., 1997), and led to the low value of SIC in the upper and middle soil. In deep soil, organic matter decomposition and root and microbial respiration release a large amount of CO₂. The CO₂ interacts with soil water to form H₂CO₃ solution, which causes the precipitation of CaCO₃ and resulting increase of SIC (Yang et al., 2010).

Aquaculture is one of the main anthropogenic activities in Jiaozhou Bay. In the process of sterilization and alkalization of aquaculture ponds, the large number of additives added into the ponds will increase the content of many different kinds of ions and nutrients in the soil and cause the salinization of the soil. This process leads to the increase in various forms of inorganic carbon, which is beneficial to the accumulation of SIC (Liu et al., 2015). Moreover, the algae resources in the aquaculture ponds are abundant, and their photosynthesis will change the physical and chemical properties of the water body, which is also one of the main factors that lead to the over saturation of calcium carbonate to produce carbonate precipitation (Obst et al., 2009). The complex and diverse management measures will affect soil SIC reserves as well. When the ponds are abandoned, the salt ions in the water accumulate on the surface of the soil under evaporation and increase the SIC reserves in the surface soil (Hren et al., 2012). Meanwhile, the biomass in aquaculture ponds is high. The decomposition activities of microorganisms can speed up the transformation of primary carbonate to secondary carbonate in surface soil, which also increases the surface SIC reserves. The leaching of water remains the main reason for the high SIC reserves in the deep soil. Additionally, related studies have shown that aquaculture activities will change the physical and chemical properties of soil in aquaculture ponds, resulting in low SIC solubility in deep sediments, which is beneficial to the retention of SIC in this layer (Quan et al., 2015).