Soil organic carbon stocks increased across the tide-induced salinity transect in restored mangrove region

Blue carbon in mangrove ecosystems contributes significantly to the global carbon cycle. However, large uncertainties maintain in the soil organic carbon (SOC) storage throughout the tide-induced salinity and alkalinity transect in the mangrove restoration region in Southern China. Total 125 soil samples were obtained to detect the SOC content and physicochemical properties. The mean SOC content of each layer ranged from 6.82 to 7.86 g kg−1, while the SOC density ranged from 2.99 to 11.41 kg m−2, increasing with soil depths. From different land covers in the study region, the SOC content varied from 4.63 to 9.71 g kg−1, increasing across the salinity and alkalinity transect, while the SOC density fluctuated from 3.01 kg m−2 in mudflats to 10.05 kg m−2 in mangrove forests. SOC concentration was favorably linked with total nitrogen (r = 0.95), and total phosphorus (r = 0.74), and negatively correlated with Cl− (r = − 0.95), electrical conductivity (r = − 0.24), and total dissolved solids (r = − 0.08). There were significant logarithmic relationships between SOC content and the concentrations of clay (r = 0.76), fine silt (r = 0.81), medium silt (r = − 0.82), and coarse silt (r = − 0.78). The spatial patterns of SOC concentration were notably affected by soil texture, physicochemical properties, and land-cover type, providing essential reference for future investigations of blue carbon budget in restored mangrove forests.

www.nature.com/scientificreports/ in Australia and the Indo-Pacific region.On a global scale, Jardine and Siikamäki 20 showed that the SOC concentration in mangroves was affected by temperature.Physical protection of carbon in soil aggregates was also an important mechanism for carbon sequestration 21 .Soil physical qualities (e.g.clay and silt concentration) were considered to protect organic materials from decomposer organisms 21 .Particles < 20 μm were essential for retaining the SOC in Baijiang soils and black soils in Jilin province 22 .In addition, a close relationship had been demonstrated among both SOC and the proportion of particle (< 2 μm) in the black soils, and a similar relationship had been found with both SOC and the proportion of fraction (2-20 μm) in the Baijiang soils 22 .These researches demonstrated the necessity of studying the SOC and the complexity of the factors controlling SOC distribution in mangrove forests.
Soil carbon in mangrove ecosystems is either autochthonous (originating from native mangrove generation) or allochthonous (originating from outside via streams) 23 .Mangrove forest is susceptible to recurrent floods and tide exposure, resulting in salt buildup and leaching.SOC stock was also highly negatively linked with tidal range 24 .Because organic carbon is less likely to be washed out of mangrove sediments under micro-tidal conditions, they can store more SOC and hence function as effective carbon sinks 25 .However, it is still unclear how tidal hydrological mechanisms and draining and soaking patterns affect carbon transmission and retention in mangrove ecosystems.
Although mangrove forests store much more carbon per unit area than in other ecological systems 26 , they are suffering from severe deforestation and face an uncertain future because of human-caused changes 5,9,[27][28][29] .Restoration programs had thus been carried out worldwide to offset for the losing of mangrove [30][31][32] .Natural and restored mangrove forests are primarily found along China's southeast shores in humid tropical and subtropical climates 33 .Zhejiang province represented the northernmost boundary of mangrove planting in China, however, little information is currently available regarding the SOC distribution in restored mangrove forests.It is urgent to carry out field observations on SOC accumulation in restored mangrove forests in South China.
We hypothesize that tide-induced salinity and alkalinity in mangrove forest have an effect on the carbon cycling and carbon deposition.By quantifying the dispersion of SOC concentration, bulk density, and SOC density in soil profiles across different land covers from lower to upper intertidal area, and investigating the relationships between SOC and soil physiochemical properties, we tested the hypotheses (1) that the SOC content increase from the low to high intertidal zone, increasing across the tide-induced salinity and alkalinity transect, and (2) the SOC concentration and distribution characteristics were significantly affected by soil texture, tidal hydrodynamic processes, and land cover.

Study area
The restored mangrove ecosystem is located at Aojiang Town, Pingyang County, Zhejiang Province, China (27°18′N, 120°30′E) (Fig. 1a,b).This area is characterized by a subtropical marine monsoon climate, with an annual mean temperature of 18.3 °C and annual precipitation of approximately 1784 mm.The lowest and highest temperatures throughout the year appear in February and August, respectively.This area is affected by continuous rainy weather, with potentially higher rainfall from March to June, and the typhoon period occurs from July to September.The tide is an irregular semi diurnal tide.The maximum tide range occurs twice in the beginning and the middle of the lunar month, and the maximum tide range can reach 603 cm.The mangrove (Kandelia obovate) was planted in 2014 on land previously occupied by Spartina alterniflora.Various land covers occurred after mangrove plantation in our study area, i.e., unvegetated mudflat (MF), mixed area of mangrove and Spartina alterniflora (MS), four mangrove communities (M1, M2, M3, M4 from onshore area to offshore area, respectively), and Spartina alterniflora (SA).

Soil sampling and analyses
We collected samples along three transects across the Aojiang estuary and seven parallel to the Aojiang River to assess the carbon stock capacity of different land cover in August 2018 (Fig. 1c).We separated the study area into seven communities with interval of 20 m, including unvegetated mudflat (MF), mixed area of mangrove and Spartina alterniflora (MS), Spartina alterniflora (SA), and four mangrove communities such as M1, M2, M3, and M4, respectively, across the tide-induced salinity and alkalinity transect from the offshore to the onshore area.Three core samples were collected and averaged for each point using an open-face stainless-steel soil sampler with an internal diameter of 3.5 cm and length of 1 m.The depth levels for sediment cores were 0-20, 20-40, 40-60, 60-80, and 80-100 cm, respectively.All sediment samples were kept in metal specimen boxes and weighed to the nearest 0.1 g to determine bulk density, after which they were packed in clean, dry, labeled plastic bags.To eliminate any pieces of wood or stones, the samples were air-dried, mixed, and filtered using a 2-mm mesh sieve.For each interval, the bulk density was computed by dividing the dry weight by the sample volume (calculated from the boxes) and soil water content, as a percentage.
The impact of environmental factors on SOC were tested according to Guo et al. 34,35 .The potassium dichromate oxidation-heating method was used to detect SOC 36 .A pHS-2C pH meter (Shanghai Youyi Instrument Co., Ltd., Shanghai, China) was used to test the pH of the soil using a soil: water (1:5) mixture.A DDS-307 conductivity meter was used to detect electronic conductivity (EC) (Shanghai Precision Instrument Co., Ltd., Shanghai, China).The concentration of total nitrogen (TN) was measured using six linked nitrogen distillers via the perchloric acid-sulfuric acid digestion method 37 .The concentration of total phosphorus (TP) was measured by the Mo-Sb colorimetric method 34 using a UVmini-1240 spectrophotometer (Shimadzu Instruments (Suzhou) Co., Ltd., China), and the concentration of total potassium (TK) was measured by acid solution-flame photometry using an FP6410 flame photometer (Shanghai Jingke Instrument Factory, Shanghai, China).Total dissolved solids (TDS) was measured by the residue drying-quality method and the Cl − concentration was measured by silver nitrate titration.According to soil particle size, the soil core was divided into clay (diameter < 2 μm), fine silt (F-Silt, 2 μm < diameter < 16 μm), medium silt (M-Silt, 16 μm < diameter < 32 μm), coarse silt (C-Silt, 32 μm < diameter < 64 μm) and sand (diameter > 64 μm).

Calculations and statistical analyses
Bulk density and SOC density were determined after removal of wood and stones using the following formulas 35 : where ρ b is bulk density (in g cm −3 ), m is the weight of wet soil in the aluminum specimen boxes (in g), V is the capacity of the aluminium specimen box (in cm −3 ), θ m is the soil water content (in %), SOCD is SOC density (in kg m −2 ), SOC i is SOC content (in g kg −1 ), and D i is depth (in cm).
Principal components analysis (PCA) is a popular multivariate technique for extracting important information from a dataset composed of several inter-correlated quantitative dependent variables 38 .Pearson's correlation coefficients have been utilized to define the degree and orientation of the linear association between two variables 39 and were presented using a correlation matrix heatmap describing the linear relationship between two parameters.The variables analyzed by PCA and the coefficient matrix heat map were SOC, pH, EC, TDS, Cl − , TN, TP, TK, clay, F-silt, M-silt, C-silt, and sand.All analyses and figures were performed using Python version 3.8.Across different land cover species, SOC, pH, EC, TDS, Cl − , TN, TP, TK, clay, F-silt, M-silt, C-silt, and sand were tested using one-way ANOVA which was performed using SPSS 19.0 (SPSS for Windows, SPSS, Inc.).Differences were regarded as significant when P was less than 0.05.ANOVA analysis was carried out for the mean values ± standard deviation of soil properties, however, no significant difference was found in the results (P > 0.05).
The SOC content in the four mangrove forest regions (M1, M2, M3 and M4) was higher at depths down to 40 cm and then decreased with increasing depth (Fig. 2a), with consistent variations in EC, TDS concentration, and Cl − (Fig. 2b-d).The highest SOC contents were found at a depth of 20-40 cm, in the order M4 (10.12 g kg −1 ) > M3 (9.24 g kg −1 ) > M2 (8.79 g kg −1 ) > M1 (7.77 g kg −1 ), showing significant difference (P < 0.05).The SOC concentration in MS was somewhat greater at 40-60 cm (6.40 g kg −1 ) than at the other layers examined, while the highest SOC in MF (4.99 g kg −1 ) was recorded in 60-80 cm.The SOC in SA declined by about 44% from 0 to 20 cm (12.50 g kg −1 ) to 60-80 cm (7.04 g kg −1 ), and the EC decreased from 2.25 ms cm −1 at 0-20 cm soil depth to 2.02 ms cm −1 at 20-40 cm, and then increased with soil depth (Fig. 2b).The EC increased slightly in the upper layers (0-60 cm) in M1 and M2, but decreased slightly in the same layers in MF and M3.The tendency of pH in M1, M2 and M3 is consistent, showing a trend of first decreasing and then increasing with increasing depth (Fig. 2e).The concentrations of TN in M4 and SA decreased markedly (P < 0.05) from 1.27 and 1.32 g kg −1 , respectively, near the surface to 0.97 g kg −1 and 0.91 g kg −1 at 0-80 cm (Fig. 2f).Interestingly, the lowest value of TN in MF and the highest value in MS were both found at 40-60 cm.The concentrations of TP in all seven landcover types decreased from 20-40 to 40-60 m soil layers (Fig. 2g).The concentrations of TK increased from 0-20 to 20-40 m soil layers except for mixed area of mangrove and Spartina alterniflora.The concentrations of TK in mixed area of mangrove and Spartina alterniflora decreased with depth within soil layer of 0-80 cm (Fig. 2h).

SOC contents in horizontal transect across different land-cover types
The SOC content increased with tidal gradient from lower to upper intertidal area, with a slight drop from M4 (9.73 g kg −1 ) to SA (9.36 g kg −1 ) (Fig. 3a).The mean SOC content in the seven land-cover types for the entire soil layer (0-100 cm) decreased in following order: M4 (9.73 g kg −1 ) > SA (9.36 g kg −1 ) > M3 (8.42 g kg −1 ) > M2 (7.77 g kg −1 ) > M1 (7.28 g kg −1 ) > MS (6.00 g kg −1 ) > MF (4.63 g kg −1 ).The differences in mean SOC contents between MF and MS, and M3 and M4 were 1.37 g kg −1 and 1.29 g kg −1 , respectively.Compared with MF, the restored mangrove and sparsely scattered Spartina alterniflora increased SOC significantly (P < 0.05).Notably, the SOC content at 0-40 cm depth layer increased in the order of MF < MS < M1 < M2 < M3 < M4 < SA, but it changed at 40-100 cm depth layer due to the drop of SOC content in SA (Figs. 2a and 3a).Soil pH of the sampling points was alkaline (pH > 7, Fig. 3b).Bulk density varied from 1.03 to 2.33 g cm −3 , with lower bulk densities in MF, M3, and M4 compared with the other stands (Fig. 3c).The concentrations of clay increased (Fig. 3d) from lower to upper intertidal area while the concentration of C-silt (Fig. 3e) and M-silt (Fig. 3f) decreased, especially in the low intertidal zone.EC (Fig. 3g), Cl − (Fig. 3h), and TDS (Fig. 3i) all increased and then decreased from seaward to landward, with the lowest value in SA.TN (Fig. 3j), TP (Fig. 3k), and TK (Fig. 3l) all increased from the low to high intertidal zone, but decreased slightly in M2 and SA.
Comparing with unvegetated mudflat, the colonization of mangrove can increase the SOC content within 60 cm and total nitrogen contents within 40 cm.No significant differences existed between S. apetala and native mangrove communities for the SOC, TN and TP contents (P > 0.05).The restoration of mangrove forests enhanced the soil carbon stock relative to mudflat but significant differences existed only between native mangrove and mudflat.Planted mangroves play important roles in enhancing carbon sequestration and nutrient storage.The mixing native mangrove community should be recommended to enlarge the area of mangrove and enhance the carbon stock capability in the future.
across different land-cover types from mudflats (MF) to Spartina alterniflora (SA).Different letters from each line indicate significant differences (p < 0.05) among different land covers according to the one-way ANOVA, which means that if the letters of any two species in the same row are the same, the difference between the index of these two land covers is not significant.

SOC stocks in restored mangrove forests
SOC density showed an increasing trend with increasing soil depth across seven land-cover types, apart from little variation in SA at 20-80 cm soil layer (Fig. 4).Similarly, there was only a slight increase between 40 and 80 cm for MS.The SOC densities ranged from 1.08 to 4.93 kg m −2 in MF, 3.05-12.53kg m −2 in MS, 3.36-16.53kg m −2 in M1, 3.63-13.39kg m −2 in M2, 1.90-9.24kg m −2 in M3, 2.49-9.76kg m −2 in M4, and 5.37-13.53kg m −2 in SA.The mean SOC densities for the seven land-cover types throughout the whole soil depth (0-100 cm) were as follow Table 1.Mean values ± standard deviation of soil organic carbon (SOC), soil pH, total nitrogen (TN), total phosphorus (TP), total potassium (TK), bulk density, electronic conductivity (EC), the concentration of Cl − , total dissolved solids (TDS), clay, fine silt (F-silt), medium silt (M-silt), coarse silt (C-silt) and sand at different soil depths.

PCA analysis of SOC and physiochemical properties
PCA captured 94.4% of variations in the first two principal components, with 73.7% explained by PC1 and 20.7% by PC2 (Fig. 5).PC1 was mainly characterized by positive loadings of C-silt and M-silt and negative loadings of SOC, F-silt, clay, and TK.PC2 was mainly characterized by positive loading of SOC and negative loadings of EC, TDS, Cl − , and TK.The SOC content was favorably associated with the concentrations of F-silt, TN, TK, and clay, and negatively correlated with the concentrations of M-silt and C-silt, according to the PCA score plot (Fig. 5).SOC content showed strong negative relationships with M-silt and C-silt (Pearson's correlation coefficient − 0.88 to − 0.86) (Fig. 6), strong positive relationships with clay, F-silt, TN, and TP (Pearson's correlation coefficient 0.74 to 0.87), and negative correlations with sand, EC, TDS, Cl − , and pH (Pearson's correlation coefficient − 0.32 to − 0.08).The concentrations of EC, TDS, and Cl − showed strong positive correlations with each other (Pearson's correlation coefficient 0.97 to 0.99), and the concentrations of clay and F-silt were strongly negatively related to M-silt and C-silt (Pearson's correlation coefficient from − 0.99 to − 0.94).We also found strong positive logarithmic relationships between SOC content and clay and F-silt, and strong negative logarithmic relationships between SOC content and M-silt and C-silt (Fig. 7).The r values between SOC content and C-silt, M-silt, F-silt, clay were 0.78, 0.82, 0.81, and 0.76, respectively.

Comparisons of SOC content with previous studies
The SOC contents (4.63-9.71g kg −1 ) in this study were similar to those in the other studies 15,40,41 (Table 2), but lower than those in Dongzhaigang (14.89-20.89g kg −1 ) 42 , the Jiulong River estuary (15.01-22.25 g kg −1 ) 43 , Vietnam (0.8-21.8 g kg −1 ) 44 , and the southern Saudi Arabian Red Sea coast (28.1-29.3g kg −1 ) 28 at lower latitudes.The SOC density in this study (2.99-11.41kg m −2 ) was comparable to previously reported values (2.6-8.5 kg m −2 ) in the same or similar depth 15,42 , but lower than in the Jiulong River estuary (13.0-17.0kg m −2 ) 43 .These differences may have been due to differences in biomass associated with climatic conditions, mangrove species, and human activities 45 .Some studies demonstrated that climatic factors were the main determinants of carbon density in mangrove forests 46 .The assessment of SOC concentration may be influenced by soil bulk density, which is an indication of the density of soil 28 .In this study, the mean soil bulk density spanned from 1.55 to 1.86 g cm −3 at 0-100 cm soil layer, which was comparable to levels in the Egyptian Red Sea coast (1.40-1.72 g cm −3 ) and the southern Saudi Arabian Red Sea coast (1.5-1.8 g cm −3 ), but higher than those in the Jiulong River estuary (0.82-0.88 g cm −3 ), Leizhou Peninsula (0.93-1.11 g cm −3 ), Dongzhaigang (1.08-1.51g cm −3 ), Futian Bay (0.93 g cm −3 , 6yrs-K.obovata and 0.79 g cm −3 , 40 years-K.obovata) 9 and Sundarbans (0.59-0.79 g cm −3 ).We found a negative association among SOC concentration and soil bulk density, which is similar with prior research 47,48 .This negative correlation demonstrated that soil bulk density has an impact on soil porosity, aeration, and construction, which in turn has an impact on SOC content.
In restored mangrove forests, soil carbon sequestration improves with age 49 .Previous studies reported that mature mangrove forests accumulated far more carbon than younger forests 4,45 .The SOC density of 4-yrs restored mangrove forest is 2.99-11.41kg m −2 in this research.Ren, Chen 50 demonstrated that after 4, 5, 8, and 10 years of restoration, SOC storage rose markedly by 3, 68, 274, and 350 kg m −2 , respectively, in four Sonneratia apetala plantations of different ages in southern China.When compared to the earlier mangroves (17 and 35 years), youthful mangroves (13 years) hold half the carbon sinks rates of older mangroves 49 .Thus, when assessing the potential of mangrove ecosystems to trap carbon, it is crucial to consider their age of the mangrove forests 9,51,52 .Furthermore, short-term mangrove recovery merely increased the top soil carbon pool, with no impact on the subsoil 53 .This information is valuable for carbon offset projects since this affects the carbon retention capacity of freshly planted mangroves and indicates that data from mature mangrove forests cannot be used as direct estimates for carbon dynamics in restored or regenerated mangroves 49 .

Relationships between SOC and soil physiochemical properties
Several physicochemical properties were correlated with SOC content in this study, indicating that soil texture is vital in SOC accretion.Previous research has found a negative association of SOC content with soil salinity 54,55 , with increased salinity accelerating microbial breakdown, limiting plant growth, and further reducing carbon accumulation 54,56 .This research also discovered a negative association of SOC concentration with salinity.In addition, electronic conductivity and the Cl − concentration were also negatively correlated with SOC content.Although SOC was not significantly correlated with pH in this study, the possible influence of pH should not be ignored; pH could affect the decomposition rate of SOC by inhibiting microbial activity 42,57 , given that most microorganisms prefer to be active when the pH value is around neutral 58 .
The logarithmic relationships between SOC content and particle size are comparable with the study reported by Hassink, Whitmore 59 , demonstrating the proportion of particles < 20 μm is intimately linked to the accumulated carbon content in this proportion in the upper 10 cm.One of the key mechanisms essential to physically protecting SOC is its propensity to bond with clay and silt particles 60 , which represent the most significant carbon sink 61 .Most of the SOC (79%-91%) was thus accumulated in the clay (< 2 μm) and silt (2-20 μm) segments 62 .Fine-textured soils contain more SOC than coarse-textured soils of the same substance source 60,63 .The quantities of carbon affiliated with the silt and clay fractions had been closely correlated with percentages of soil particles in these fragments in both temperate and tropical regions 60 , suggesting that finer sediments may provide a more reactive area that can gather organic carbon, thus protecting organic carbon from remineralization 44 .Besides, mangrove forests have been proved to effectively increase the clay content in soil, which is related to higher carbon accumulation 64 .
Positive relationships were observed in the concentrations of SOC, TN, and TP in the restored mangrove ecosystems (Fig. 5, 6).Mangrove forests, growing in eutrophic waters, have the potential to accumulate SOC, TN, and TP at higher rates than mud flat in this study.Spartina alterniflora enhanced the SOC, TN, and TP contents significantly (P < 0.05) comparing with mud flat and mangrove ecosystems (Fig. 3).Both mangrove forests and Spartina alterniflora could uptake a tidal nitrogen and phosphorus nutrients to satisfy its requirement for growth (Fig. 3).Previous studies indicated that Spartina alterniflora could increase SOC and TN contents by 0.37 and 7.43

Variations in mangrove SOC across tide and salinity gradients
The SOC content was obviously higher in mangrove forests (7.28-9.73g kg −1 ) than in the adjacent mudflats (4.63 g kg −1 ), and was higher in landward compared with seaward communities in this study.This result may be attributable to less erosion and a more stable hydrological environment in the landward communities, thus allowing more carbon to be stored in the soil 13 .Most of the soil carbon stock in the top meter is accounted for by autochthonous carbon, such as litter from plants and roots 6,69 , while some is derived from the tide 69,70 .Similar results were found in estuarine mangrove communities in South China 12 and in the mangroves at Dongzhai Harbor National Natural Reserve 69 .Furthermore, SOC content was substantially greater near the estuary than in the more salinized coastal areas 71 , which is in agreement with our result that SOC content was negatively correlated with TDS.Previous study showed that the SOC in the juvenile mangrove areas is mostly of allochthonous origin (estimated on average at 79%) 72 .And the mangrove areas that are about 50 m from the beach and are not inundated daily.This may explain why the SOC density increased from 3.01 to 10.05 kg m −2 , and then decreased to 5.15 kg m −2 with the tidal gradient from lower to upper intertidal area.This is consistent with the fact that organic carbon levels steadily grow in surficial layers from the seashore up to a certain extent and then fall 24 .The SOC content was greater in SA than other land-cover categories, except for M4, which might have been due to the plant species as well as the tide/salinity gradient.Interestingly, Lunstrum and Chen 9 reported no significant change in SOC concentrations between S. apetala and K. obovata over a 6-year period.Further studies are therefore needed to investigate the aboveground and subsurface biomasses of mangrove communities at different locations and in forests of different ages in Zhejiang.

Conclusions
Our findings provide compelling evidence that restored mangrove forests have superior capacity to sequester carbon in the sediments compared with the mudflat.In the region of mixed mangrove and Spartina alterniflora (MS) and single mangrove ecosystems, the SOC content increased initially and subsequently dropped with increasing soil depths.The SOC density rose as soil depths increased.A pattern of increasing SOC content (4.63-9.73g kg −1 ) along the tide-induced salinity transect was found, while SOC density varied from 3.01 kg m −2 on mudflats to 10.05 kg m −2 in mangrove forests.SOC content was shown to be favorably connected with TN, TP, and negatively associated with Cl − , EC, and TDS.SOC concentration has strong logarithmic associations with clay, F-silt, M-silt, and C-silt.More extensive research is required to increase our comprehension of carbon sinks and its regulating www.nature.com/scientificreports/mechanisms in mangrove ecosystems.To meet its vision to support carbon neutrality by 2060, China might also boost scientific studies on coastal habitats, safeguard the functional and structural stability of existing coastal wetlands, prevent catastrophic coastal wetlands activities, reestablish as well as strengthen the "blue carbon" role.

Figure 4 .
Figure 4. SOC density according to land-cover type at soil depths of 0-100 cm.Five colors from the bottom up represent 0-20 cm, 20-40 cm,40-60 cm, 60-80 cm, and 80-100 cm.Values (in kg m −2 ) above the column represent the average SOC density among five depth layers for different land types.

Figure 6 .
Figure 6.Correlation matrix heatmap generated in Python 3.8 showing Pearson's correlation coefficients for soil physicochemical properties.Coefficients ranged from − 1 to 1, with − 1 indicating a negative relationship between two variables, 1 indicating a positive relationship, and 0 indicating no relationship.