Composition and Biogeochemical Effects of Carbohydrates in Aerosols in Coastal Environment

Abstract

We adopted a simple and rapid measurement method to analyze the concentrations of monosaccharides (MCHO) and polysaccharides (PCHO) in carbohydrates, a subset of organic carbon found in size-fractionated atmospheric particles. Seasonal and source-related factors influenced carbohydrate concentrations, with total water-soluble carbohydrates (TCHO) accounting for approximately 23% of the water-soluble organic carbon (WSOC) in spring when biological activity was high. We observed that the mode of aerosol transport significantly influenced the particle size distribution of carbohydrates, with MCHO exhibiting relatively high concentrations in fine particles (<1 μm) and PCHO showing higher concentrations in coarse particles (>1 μm). Moreover, our results revealed that MCHO and PCHO contributed 51% and 49%, respectively, to the TCHO concentration. This contribution varied by approximately ±19% depending on the season, suggesting the importance of both MCHO and PCHO. Additionally, through the combined use of principal component analysis (PCA) and positive matrix factorization (PMF), we determined that biomass burning accounts for 30% of the aerosol. Notably, biomass burning accounts for approximately 52% of the WSOC flux, with MCHO accounting for approximately 78% of the carbon from this source, indicating the substantial influence of biomass burning on aerosol composition. The average concentration of TCHO/WSOC in the atmosphere was approximately 18%, similar to the marine environment, reflecting the relationship between the biogeochemical cycles of the two environments. Finally, the fluxes of MCHO and PCHO were 1.10 and 5.28 mg C m⁻² yr⁻¹, respectively. We also found that the contribution of atmospheric deposition to marine primary productivity in winter was 15 times greater than that in summer, indicating that atmospheric deposition had a significant impact on marine ecosystems during nutrient-poor seasons. Additionally, we discovered that WSOC accounts for approximately 62% of the dissolved organic carbon (DOC) in the Min River, suggesting that atmospheric deposition could be a major source of organic carbon in the region.

1. Introduction

Organic aerosols constitute a significant portion of atmospheric particulate matter, influencing regional and global climate dynamics [1]. The International Energy Agency (IEA) estimated that the carbon flux emitted into the atmosphere in 2022 was 36,800 Tg C yr⁻¹, with organic carbon (OC) representing 30% of these emissions [2]. OC can be further categorized into water-soluble organic carbon (WSOC) and water-insoluble organic carbon (WIOC) [3], with WSOC typically constituting 10% to 80% of total organic carbon (TOC) [4,5,6,7].

Carbohydrates, as part of WSOC, are commonly found in aerosols from various environments, including those from rural, urban, and marine areas [8,9]. These compounds, which serve as the primary energy source for living organisms, are predominantly introduced into the atmosphere through atmospheric deposition [10], and are sourced from natural emissions (e.g., plants, fungal spores, soil, and dust resuspension) and anthropogenic activities like industrial emissions, and agricultural activities, particularly biomass burning. Due to their universality and abundance, carbohydrates can be used to clarify the sources and transport of atmospheric organic aerosols [11], identify their sources and assess their atmospheric transport by serving as tracers for organic molecules [12,13,14,15].

Increases in anthropogenic pollutant emissions [16], changes in land use [17], and elevated emissions of biogenic particles [1] have resulted in the long-range transport of these pollutants to the North Pacific and even the American continent [18,19].

PCA is a data analysis method aimed at explaining the maximum variability in an original dataset using the fewest principal components (factors) [20]. Although it does not account for the temporal variability in pollutants, it can be used to assess their correlation with the overall variation in aerosol composition [21]. On the other hand, PMF is a source apportionment modeling method primarily used to deconvolute the sources of pollutants in atmospheric aerosol samples [22]. PMF can be used to identify and quantify contributions from different pollution sources and is commonly employed to study the composition and sources of atmospheric aerosols.

In summary, in this study, we focused on analyzing the chemical composition of aerosols in Matsu, which is located in the southern East China Sea. By employing numerical analysis models such as principal component analysis (PCA) and positive matrix factorization (PMF), we aimed to identify aerosol sources, assess their contributions, and evaluate the stability and reliability of the results.

2. Materials and Methods

2.1. Aerosol Sampling

The atmospheric sampling station used in this study was located on the Matsu Islands (26.167° N, 119.917° E) at the mouth of the Min River, adjacent to mainland China to the west (approximately 97 km) and the Taiwan Strait to the east, 220 km from Taiwan (Figure 1). The climate of the study area is a subtropical oceanic climate, with the sampling area primarily influenced by the northeast monsoon from winter to spring, and the southwest monsoon from summer to autumn. The sampling site was approximately 10 m from the coastline and had an elevation of 10 m above sea level to avoid interference from human influences. The sampling period in this study was from April 2019 to September 2020 (108 samples in total).

We used the high-flow air samplers (Tisch TE-5170: Tisch Environment, Inc., Cleves, OH, USA), combined with a five-stage high-volume cascade impactor (TE-235) and porous quartz filters (TE-230-QZ, Tisch) for sampling. The impactors had cutoff diameters equivalent to 50% efficiency, and they were used for the collection of size-fractionated particles as follows: stage 1 (>7.2 µm), stage 2 (3.0–7.2 µm), stage 3 (1.5–3.0 µm), stage 4 (0.95–1.5 µm), and stage 5 (0.49–0.95 µm). Particles that passed through the CI were trapped on a backup filter (<0.49 µm). The flow rate was set to 0.85 m³ min⁻¹, and sampling was conducted continuously for 168 h.

After sampling was completed, the filter samples were stored at −24 °C until analysis. Each filter obtained from aerosol sampling was extracted by sonication with 200 W of output power (DC200H; Delta Electronics, Inc., Taipei, Taiwan) for 3 h in 100 mL of Milli-Q water (>18 MΩ cm). After extraction, the extracts were immediately analyzed for various chemical species. Complete details of the dry deposition sampling and extraction procedures have been described previously in Chen and Chen [23].

2.2. Species Analysis

In this study, the analysis of carbohydrates was divided into two parts: total water-soluble carbohydrates (TCHO) and monosaccharides (MCHO). TCHO was measured according to the phenol–sulfuric acid method as described by DuBois, Gilles [24]. The principle of this method is dehydration of polysaccharides (PCHO) by concentrated sulfuric acid to form uronic acid and hydroxyurea formaldehyde. Subsequently, these compounds undergo condensation with phenol to produce an orange–red-colored compound. Measurement was performed using a colorimetric method, and the absorbance values were measured at a wavelength of 490 nm using a spectrophotometer. The reaction formula is shown in Figure 2.

The determination of MCHO was based on the TPTZ (2,4,6-tripyridyl-s-triazine) reduction method, following Myklestad, Skanoy [25]. The principle for the method is the oxidation reduction reaction of dissolved sugars in water or nonreducing sugars, which are hydrolyzed to reduce polysaccharides under alkaline conditions. During this process, Fe³⁺ ions in the reagent are reduced to Fe²⁺, forming a blue–purple-colored compound with TPTZ. Measurement was conducted using a colorimetric method, and the absorbance values were measured at a wavelength of 595 nm using a spectrophotometer. The reaction equation is as follows:

RCHO + 2Fe (CN)₆³⁻ + 3OH⁻ → RCO₂⁻ + 2Fe (CN)₆⁴⁻ + 2H₂O

(1)

2 Fe (CN)₆⁴⁻ + TPTZ → Blue-Violet compound

(2)

In addition to the analysis of carbohydrate species, analyses of carbon, nitrogen, phosphorus species, and major ions (Na⁺, K⁺, Mg²⁺, Ca²⁺, Cl⁻, and SO₄²⁻) were conducted. The water-soluble nitrogen species were divided into water-soluble organic nitrogen and water-soluble inorganic nitrogen. The inorganic nitrogen content was analyzed by methods in Pai, Yang [26], Pai and Riley [27], and Pai, Tsau [28]. Water-soluble inorganic phosphorus (WSIP) was measured following the molybdate blue method proposed by Murphy and Riley [29], and water-soluble organic phosphorus (WSOP) was measured as described by Huang and Zhang [30]. The concentrations of major ions were determined using an ion chromatograph (ICS-1000, Dionex, Sunnyvale, CA, USA) via chromatographic separation of water-soluble anions (Cl⁻, SO₄²⁻) and cations (Na⁺, K⁺, Mg²⁺, Ca²⁺) in the samples.

2.3. PCA and PMF Models

In this study, we employed principal component analysis (PCA) and positive matrix factorization (PMF) to analyze the sources of individual species. PMF is more suitable for determining the sources of pollutants and their temporal variability without considering their correlations. PCA and PMF are both factor analysis methods, but the second is especially suitable for environmental data due to its nonnegativity constraint [21,31].

PCA is a multivariate analysis method [32], and the analysis was performed using the SPSS model (IBM SPSS Statistics, version 19.0). Its purpose is to transform many correlated variables into new independent variables, which are linear combinations of the original variables. This simplifies the complexity of variable analysis and facilitates further discussion of the relationships between the new variables. The formula for PCA is as follows:

$$X_{centerd}=X-\mu$$

(3)

The mean is calculated, where X_centerd represents the data after mean subtraction; X denotes the original data; and μ stands for its feature mean.

$${\mathrm{C}}_{ij}={\displaystyle \frac{1}{n}}{X}_{centered}^T{X}_{centered}$$

(4)

The covariance matrix was computed, eigen decomposition was performed, the eigenvector projections of the principal components were retained, and in the results of principal component analysis were obtained.

PMF is a statistical model for matrix factorization [33], and the analysis was carried out in EPA PMF 5.0 developed by the U.S. Environmental Protection Agency. It is employed to analyze potential pollution sources in the atmosphere and serves as a receptor model in air quality modeling. Based on the composition profile of pollutants from pollution sources, PMF identifies and quantifies the sources of air pollutants at receptor sites using mathematical and statistical procedures. The principle involves computing the chemical component errors in particulate matter using weighting factors and subsequently determining the major pollution sources and their contributions through the minimum squares method. The explained variation (EV) is defined as follows:

$${EV}_{kj}={\displaystyle \frac{\sum _{i-1}^P\left|G_{ik}{F}_{ij}\right|/S_{ij}}{\sum _{i=1}^n(\sum _{i-1}^P\left|G_{ih}{F}_{hj}\right|+\left|E_{ij}\right|)/S_{ij}}}$$

(5)

When EV_kj approaches 1, chemical component (j) represents factor (k) more prominently.

2.4. Air Mass Backward Trajectories

We utilized the Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT Model) [34] provided by the Air Resources Laboratory (ARL) of the National Oceanic and Atmospheric Administration (NOAA). Data analysis was conducted using the TrajStat software developed by Meteo-Info [35]. In this software, the principle of the analysis is based on geographic information system (GIS), and cluster analysis is employed to categorize air mass trajectories into four different directions. This approach aids in estimating the transport pathways of airflow reaching the study area and identifying the sources of air masses.

3. Results

3.1. Background Information

The Matsu Islands are influenced by the Asian monsoon throughout the year. In this study, spring is defined as March to May, summer as June to August, autumn as September to November, and winter as December to February. Backward trajectories are set to 168 h with a height of 500 m above the mixed layer height. The air mass trajectories during the sampling period are traced back based on the seasons to estimate the transport pathways of air masses reaching the study area (Figure 3).

The results indicate that air masses originating from the northeastern regions of China and Northeast Asia are transported to the study area by the northeasterly monsoon, predominantly during autumn to spring (September to April) when sampling occurs. During summer (June to August), the prevailing southwestern monsoon brings air masses mainly from nearby seas and the South China Sea region, with minor contributions from South Asia and Southeast Asian countries.

The studies of Fu et al. [9] and Zhu et al. [36] indicated that the seasonal distribution of carbohydrate concentrations varies by region, influenced by factors such as temperature and rainfall. As shown in Figure 4 below, the highest carbohydrate concentrations were observed in spring, while peak temperatures occurred in summer. Rainfall, primarily influenced by typhoons, peaked during summer 2019 and autumn 2020. Our analysis indicates that there is no clear direct relationship between local temperature, rainfall, and carbohydrate concentrations. This suggests that aerosol concentrations in the region are primarily influenced by sources from the nearby continent.

3.2. Concentration and Distribution Characteristics

During the sampling period, the average mass concentration of total aerosols was 40.0 ± 7.0 μg m⁻³, with coarse particles averaging 18.9 ± 4.4 μg m⁻³ and fine particles averaging 21.1 ± 5.2 μg m⁻³. The highest mass concentration observed in April 2020 was 52.5 μg m⁻³, likely due to vigorous springtime biological activities. From winter to spring (December 2019 to May 2020), the mass concentration of fine particles accounted for more than 50% of the total mass due to the influence of continental sources from the northeasterly monsoon. The highest proportion of fine particles occurred in April 2019 (64%). During summer, the southwestern monsoon brought a significant amount of sea-derived coarse particles, leading to a greater proportion of coarse particles.

The percentages of all the analyzed aerosol particles are presented in Figure 5, with WSTC accounting for the largest portion, which can be further divided into water-soluble inorganic carbon (WSIC) and organic carbon. Among them, WSOC accounts for approximately 75% of the WSTC, and sulfate ions are the second most abundant species; hence, it can be inferred that the primary sources of local aerosols may be influenced by these two main substances, namely biomass burning and anthropogenic emissions from fossil fuel combustion.

The monthly average concentration of TCHO during the sampling period was 160 ± 47.7 ng m⁻³, with the highest value occurring in April 2020 at a concentration of 254 ng m⁻³, as shown in Figure 6. The concentrations were greater during spring and winter, exhibiting seasonal variations. This may be attributed to a peak in agricultural activities and heightened biological processes during the spring season. The organic substances present in these seasons could react with other pollutants, forming secondary aerosols or photochemical pollutants, and thereby deteriorating air quality [37]. To further investigate aerosol composition, it is essential to analyze the concentration and types of components in different aerosol particle size fractions.

3.3. Particle Size Distribution and Seasonal Variations

In the study by Chen and Chen [38], the ratio of nutrients in coarse to fine particles (C/F) may reflect the transport mechanism or source of aerosols. In this study, the C/F ratios of MCHO and PCHO were 0.09 and 1.01, respectively. The C/F ratio of PCHO was much greater than that of MCHO, indicating that MCHO may be related to fine particles. In the particle diameter distribution chart (Figure 7), it can be observed that MCHO was mainly in fine particles, especially in the last two stages (<0.95 μm), while PCHO was mainly distributed on coarse particles (>3 μm). Different types of saccharides may originate from different sources, and these sources may be associated with specific particle sizes of pollutants. It is speculated that this is related to selective adsorption by aerosol particles. MCHO typically have smaller molecular sizes and larger surface areas; as a result, it is more easily adsorbed by fine particle filters. Conversely, PCHO is usually larger and therefore more likely to adhere to coarse particles.

In May 2020, when the biological effects were strong, the proportion of TCHO/WSOC was approximately 23%. These results are consistent with the research of Mullaugh, Byrd [39], who reported that carbohydrates typically constitute less than 1% of the WSOC. Nevertheless, during periods of increased pollen or localized fires, this percentage can increase to 10–35%, similar to our observations. The average proportion of TCHO/WSOC is also close to the proportion mentioned in the study of Hung, Tang [40] (TCHO/DOC is approximately 18%), indicating a correlation between the atmosphere and ocean biogeochemical cycles.

The average concentration of MCHO was 13.6 ± 18.8 ng m⁻³, accounting for approximately 51% of the TCHO change seasonally, as shown in Figure 8. In summer, the PCHO abundance was much greater than the MCHO abundance, which was attributed to the fact that PCHO was present primarily on coarse particles. The transport modes of these coarse particles resemble those of marine particles via the summer southwest monsoon; similarly, during the winter season, which is dominated by the northeast monsoon, these coarse particles are affected by fine suspended particles from China [41,42].

Finally, according to the organized global carbohydrate data shown in Figure 9, the TCHO and MCHO concentrations were the lowest in the western Pacific, indicating the crucial influence of carbohydrate concentration on aerosol transportation modes and sources [1]. At the Finokalia station in the Eastern Mediterranean, MCHO (glucose and levoglucosan) was the most abundant saccharide. The concentrations of saccharides were greater in spring and lower in autumn, showing seasonal variations [43]. Carbohydrates at the Birkenes station in Norway are heavily influenced by wood combustion and reach their peak values in summer and autumn [8]. In Howland, Maine, USA, the highest total saccharide concentration was observed during prevalent wildfire smoke events, with MCHO accounting for 40–75% of the total relative saccharide abundance during the sampling period [13]. On Chichi Jima, which is an island in the Pacific, the highest saccharide concentrations occur in winter and spring, primarily due to biomass burning, but are affected by monsoon factors [1]. In Seoul, South Korea, the average total saccharide concentration is highest in winter and increases from spring to summer, mirroring the agricultural season in Asia, as Seoul is surrounded by the major agricultural regions of Korea [44]. Similarly, in Beijing and Baoji, China, saccharide compound concentrations peak in winter, suggesting the significant impacts of secondary pollution sources and seasonal variations [45,46]. Finally, Jeju Island in the northern East China Sea, Okinawa Island in the western East China Sea, and the area in this study were compared. Spring was the main season for saccharide abundance, with the highest concentrations found in Okinawa, primarily due to emissions from local broad-leaved forests [36]. Consequently, this study suggested that WSOC abundance was greater in these regions than in other regions, which might be due to anthropogenic emissions from China.

3.4. Source Analysis by PCA and PMF

We analyzed the water-soluble anions and cations in the samples. The most common water-soluble ions in the ocean are Na⁺, Cl⁻, and Mg²⁺, which are typically regarded as sources of sea salt in aerosol particles [23,47]. The concentration of non-sea-salt (nss) ions can be calculated using the equation proposed by [38]. The equation is as follows:

[nss-SO₄²⁻] = [SO₄²⁻] − 0.06 × [Na⁺]

(6)

[nss-K⁺] = [K⁺] − 0.02 × [Na⁺]

(7)

[nss-Ca²⁺] = [Ca²⁺] − 0.02 × [Na⁺]

(8)

Nss-SO₄²⁻ is primarily contributed by fuel combustion and vehicle emissions [48]; nss-K⁺ serves as a typical tracer for biomass burning [49]; and nss-Ca²⁺ is predominantly governed by mineral and soil weathering [50]. These non-sea-salt ions typically exhibit higher concentrations in areas influenced by human activities and in terrestrial particles.

Carbohydrates, which represent primary emissions, are considered to be tracers for biomass burning [51,52]. The correlation analysis results between TCHO, MCHO, and nss-K⁺ show p-values of less than 0.001, indicating statistical significance. The regression coefficients (R²) range from 0.4 to 0.7, suggesting a moderate correlation. This indicates that in this area, saccharides are not only indicative of biomass burning but also have other sources.

According to the studies by [53,54], the combined use of PCA and PMF can provide a better understanding of aerosol sources. Therefore, this study also employs these two methods for analysis.

PCA and PMF were used for coarse and fine particles, respectively. In the case of coarse particles, PMF separated anthropogenic emissions and biomass burning sources based on one factor according to the PCA results and did not identify the sources of the fourth factor. However, marine sources were successfully distinguished. In the analysis between fine particles and coarse particles, anthropogenic emissions and biomass burning sources were separated, and successful identification of sources from the crust and fourth source was achieved.

We found that the results for different particle sizes using PCA and PMF methods were not entirely consistent. The discrepancy may stem from the different interpretations of interactions between components in the two methods. Additionally, splitting the original data into smaller subsets results in a reduction in the number of data points within each subset, leading to increased data sparsity [55]. Unlike PMF, PCA tends to condense components into fewer principal factors, simplify the data and enhance interpretability, but the resulting data sparsity may lead to model instability. Therefore, the analysis results for the total concentrations of each species are more reliable than those for coarse and fine particles.

Table 1. PCA for each species.

	Factor 1	Factor 2	Factor 3	Factor 4
NO3−	−0.04	0.65	0.59	0.18
NH4+	0.90	−0.11	0.27	−0.04
WSIN	0.73	0.25	0.43	0.06
WSON	0.31	0.28	0.61	0.20
WSIP	0.22	0.13	−0.27	0.63
WSOP	0.30	0.50	−0.41	0.51
TCHO	0.78	0.30	−0.10	−0.11
MCHO	0.91	−0.17	−0.03	−0.17
PCHO	−0.27	0.69	−0.10	0.10
WSOC	0.90	−0.12	−0.14	−0.24
WSIC	0.80	−0.09	−0.23	−0.10
Na+	−0.16	0.76	−0.24	−0.41
K+	0.84	0.28	−0.27	0.08
Ca2+	−0.07	−0.09	−0.05	0.94
Mg2+	0.07	0.82	−0.20	0.05
Cl−	−0.17	0.73	−0.08	−0.52
SO42−	0.95	−0.03	0.03	−0.12
nss-K+	0.76	0.40	−0.14	0.21
nss-Ca2+	−0.34	0.74	0.29	0.00
nss-SO42−	0.88	−0.06	0.15	−0.09
Eigenvalue	7.69	4.91	1.72	1.38
Variability	38%	25%	9%	7%
Sources	Biomass Burning	Marine Source	Anthropogenic emissions	Crustal Source

presents the results of the principal component analysis for the total concentrations of each species, which revealed four influencing factors that explained more than 79% of the total variance. Only factors with eigenvalues greater than 1 are presented in the table, with smaller factors considered less significant in affecting species source distribution [38,56]. Only factors with factor loadings >0.7 for Factor 1, >0.6 for Factor 2, and >0.5 for Factors 3 and 4 are considered and are highlighted in bold.

First, Factor 1, dominated by NH₄⁺, WSIN, TCHO, MCHO, WSOC, WSIC, K⁺, SO₄²⁻, nss-K⁺, and nss-SO₄²⁻, explained 38% of the variance. Biomass burning is a significant source of atmospheric water-soluble organic compounds, and agricultural fires generate large amounts of ammonium [57,58]; thus, Factor 1 is inferred to represent biomass burning sources. Factor 2 exhibited high to moderate levels of Na⁺, Mg²⁺, Cl⁻, and nss-Ca²⁺, suggesting marine sources, and accounted for 25% of the variance. Factor 3, characterized by significant amounts of NO₃⁻ and WSON, explains approximately 9% of the variance and is classified as anthropogenic emissions resulting from human activities. Factor 4, which was enriched with substantial amounts of the crustal elements WSIP (50%) and Ca²⁺ (38%), represented continental sources from fossil fuel combustion, and explained approximately 7% of the variance.

The results of the positive matrix factorization analysis for the total concentrations of each species are shown in Figure 10. Factor 1 is dominated by NH₄⁺ (85%), MCHO (80%), and SO₄²⁻ (69%), indicating that biomass burning is a significant source of atmospheric water-soluble organic compounds. Therefore, Factor 1 is inferred to represent biomass burning sources. Factor 2 contains 51% Na⁺, 72% Cl⁻, and 52% Mg²⁺, suggesting marine sources. Factor 3 is characterized by nitrogen species derivatives: NO₃⁻ (72%), WSIN (43%), and WSON (65%), classified as anthropogenic emissions resulting from human activities. Phosphorus species in the atmosphere mainly originate from soil, minerals, dust, etc. [59] Factor 4 is enriched with substantial amounts of the crustal elements WSIP (50%) and Ca²⁺ (38%), representing continental sources from fossil fuel combustion. The similarity between the results of the two different numerical simulations and their respective species source distributions indicates a high degree of model fit, with the distinct characteristics of each species representing local aerosol sources.

3.5. Flux and Implications for Carbon Export Production

By estimating fluxes, the potential impact of atmospheric particulate matter input on marine primary productivity can be assessed. Therefore, referring to the flux formulas provided by [16,60], the flux formula is as follows:

$$F_d=C_d\times V_d$$

(9)

where F_d represents the dry deposition flux, C_d represents the concentration of the species (ng m⁻³), and V_d represents the rate of aerosol dry deposition (cm s⁻¹). The deposition rate is different for aerosol particles of different sizes, with coarse particles having a rate of 2.0 cm s⁻¹ and fine particles having a rate of 0.1 cm s⁻¹.

In our study, the annual flux of WSTC was approximately 137 mg C m⁻² yr⁻¹. In the WSTC, OC accounts for 84%, confirming that carbon in the atmosphere mainly exists in the organic form and is the primary component of suspended particles, which affects atmospheric visibility and optical properties. The source contributions of WSOC are shown in Figure 11, with biomass burning accounting for 52% and crustal sources accounting for 36%, reflecting the diversity of their sources. Although TCHO accounted for only 0.4% of the WSOC flux, it had a significant impact on ecosystems. The estimated annual fluxes of MCHO and PCHO were 1.10 and 5.28 mg C m⁻² yr⁻¹, respectively, with PCHO accounting for 83% of the TCHO, indicating their importance in carbohydrate aerosols and greater biological availability. Model analysis revealed that biomass burning contributed approximately 78% to the MCHO flux. Approximately 76% of the PCHO flux originated from the ocean. In addition to the influence of particle size and transportation mode, this may also be related to bubble rupture in sea surface microlayers (SMLs) as mentioned by Russell, Hawkins [61], which requires further research for validation. Overall, these results emphasize the importance of biomass burning and marine sources for atmospheric carbon compounds.

Primary productivity is an important concept in ecology, and refers to the rate at which photosynthetic or chemosynthetic organisms convert energy and nutrients into organic matter. We estimated the deposition of OC from the atmosphere to the ocean and combined it with the utilization rate of OC by marine plankton to infer the potential contribution of the atmosphere to oceanic primary productivity.

We considered the total flux of WSTC in aerosols as the atmospheric primary productivity, averaging 0.37 mg C m⁻² d⁻¹. Although the contribution of atmospheric primary production to marine primary production is not large, its seasonal variations have a significant impact on marine ecosystems. In winter, the primary productivity provided by the atmosphere to the ocean is more than fifteen times greater than that in summer. This indicates that during seasons of relative nutrient scarcity and lower ocean productivity, atmospheric deposition supplies additional carbon sources, maintaining the stability of the ecosystem and significantly impacting marine ecosystems. In this study, the measured flux of WSOC was 0.32 mg C m⁻² d⁻¹, which is similar to the results of Matsumoto, Kodama [62]. Zhao, Qi [63] assumed that the deposition flux obtained for the region is applicable to the entire East China Sea region, which has an area of 500,000 km². This calculation results in an annual WSOC flux for the East China Sea of 6.8 × 1010 g yr⁻¹. The Min River, a major estuary of the East China Sea located near our study area, has an annual dissolved organic carbon (DOC) flux of 11 × 1010 g yr⁻¹ [64], with atmospheric deposition accounting for approximately 62%. This highlights that atmospheric deposition of WSOC is a significant carbon source in this region and emphasizes the importance of atmospheric deposition in the carbon cycle of marginal seas.

4. Conclusions

In this study, we chose carbohydrate compounds as tracers of organic matter in aerosols to investigate their concentrations, seasonal variations, and impacts on WSOC and aerosol particles at the southern end of the East China Sea. The results showed that the highest concentrations of carbohydrate species occurred in spring, indicating that both biological activity and anthropogenic activities were the main contributors to carbon species. However, high MCHO and PCHO concentrations did not occur at the same time, suggesting a complex composition and diverse sources of carbohydrates. MCHO was mainly associated with fine particles (C/F ratio averaging 0.09), indicating sources from biomass burning or long-distance transport, while PCHO was primarily associated with fine particles (C/F ratio averaging 1.01), indicating sources from the crust or the ocean. There were also significant statistical relationships between the concentrations of different ions and particle sizes. The biomass burning indicator (nss-K⁺) and fossil fuel indicator (nss-SO₄²⁻) were significantly correlated with WSOC and MCHO.

The PCA and PMF analysis results mainly indicated four factors, with biomass burning being the largest contributor, indicating significant anthropogenic influences from nearby continents. The next most significant factor was marine sources, suggesting that waves from nearby seas and the South China Sea could be the main sources of sea salt ions. The annual average fluxes of MCHO and PCHO were 1.10 and 5.28 mg C m⁻² yr⁻¹, respectively, accounting for only 0.4% of the OC flux. MCHO was mainly derived from biomass burning, while the flux of PCHO was much greater than that of MCHO, indicating greater biological availability of PCHO. Both MCHO and PCHO played important roles in the atmosphere, and the ratio of carbohydrates to OC in the atmosphere was comparable to that in the ocean, indicating a biochemical correlation between the two. Additionally, the flux calculations indicated that atmospheric deposition significantly contributed to primary productivity, with the greatest contribution occurring in winter, which was fifteen times greater than that in summer. This had a substantial impact on the primary productivity of marine ecosystems. Furthermore, atmospheric deposition was a major source of organic carbon in the Min River estuary, demonstrating that the flux of organic carbon in the atmosphere not only affects marine primary productivity but also has a significant impact on the carbon cycle in marginal seas.