This study focuses on the selection of mangrove and non-mangrove forests in river basins located in Zhanjiang Bay (ZJB) in southern China. ZJB is a typical subtropical semi-enclosed bay with a mild climate and fertile water, surrounded by the main urban area of Zhanjiang City. The area’s topography is complex, making it relatively suitable for mangrove growth (Syahid et al., 2020). The region of ZJB is connected to the South China Sea through a narrow channel (< 2 km), which allows for the exchange of suspended seawater sediments between ZJB and the outer part of the bay (Zhou et al., 2021). Sediment samples were collected from five stations in ZJB, and detailed information for each station is shown in Figure 1 and Table 1 . In addition, to explore the MPs difference in the mangrove and non-mangrove sediments in ZJB, sediment samples (three replicates) were collected from mangrove and unvegetated bare flat sites during low tide.

The mangrove forests are in the estuarine and coast zone, which is easily impacted by the river discharge and wastewater input by human activities. Nanliu River (S1) is situated in close proximity to industrial plants, including fertilizer production plants, which transport significant quantities of treated industrial wastewater to coastal waters (Zhang et al., 2019). Lvtang River (S2) is heavily polluted due to the discharge of urban sewage. Sino-Australian Garden (S3) is a popular tourist attraction with a water surface area of 42, 400 m². The northern part of ZJB is fed by the Suixi River (S4) with an area of 1486 km². It carries runoff from major agricultural areas and is the largest freshwater flow into ZJB (Zhang et al., 2019; Zhang et al., 2020b). S5 is the Potou Primary School, which has a large population in the vicinity.

Sampling was completed on 30^th October, 2021, and all tools were cleaned with Milli-Q water before sampling. Two separate sample squares measuring 1.5 m x 1.5 m, spaced 2 m apart, were randomly set up at the sampling stations. Surface sediment (0-5 cm) was collected using a clean stainless steel grab sampler to remove debris such as gravel and shells. The collected sediment was then transferred to a clean aluminum box with markers. Each sediment sample consisted of at least1 kg of wet sediment. A total of 20 surface sediment samples (10 points × 2 replicates) were collected. It is crucial to avoid any contamination during transportation to the laboratory for research.

The MPs were identified using the method proposed by Masura et al. (2015) with modifications. The samples underwent pre-treatment, mainly for ablation and the measurement of particulate organic carbon. Following pre-treatment, the MPs were quantified and characterized. To avoid contamination, sediment samples were first spread evenly in an aluminum tray, covered with aluminum foil, and dried in an oven at 75°C. A solution of sodium chloride (450 ml, density 1.2 g/cm³) was added to a 1 L beaker containing 200 g of dried sediment. The mixture was stirred for 5 minutes and left to stand for 15 minutes until the suspended solids settled. The upper layer of liquid was passed through a 45 μm stainless steel sieve and the residue on the sieve was washed three times with ultrapure water and then transferred to a 250 ml beaker. To avoid interference to natural organic matter in the sediment, a solution of 90 ml of 30% H₂O₂ and 90 ml of 0.05 M FeSO₄ was added to the beaker for disintegration. The beaker was then heated in a water bath at 75°C for 12 hours and cooled at room temperature (25°C) for 24 hours. The samples were vacuum filtered through a 10 µm cellulose acetate filter membrane and covered with aluminum foil. Finally, the filter membranes were dried in an oven at 60°C to facilitate further analysis. Content of organic carbon of each sediment sample was estimated by loss on ignition (LOI) (Samper-Villarreal et al., 2016; Huang et al., 2021). The mean precision of the ratio of organic carbon to organic matter (LOI) was 12.0% in this study. To measure particulate organic carbon (POC), the sediment was dried in an oven at 60°C for 12 hours and weighed to determine its density after drying. The sediment was then burned in a muffle furnace at 500°C for 4 hours, and the burn-off rate was measured (Huang et al., 2021).

MPs were systematically quantified using a stereomicroscope (Nikon SMZ1270, Tokyo, Japan) at magnifications up to 40 × 40 for MPs ranging from 0.45 to 5 mm (Masura et al., 2015). The following criteria must be met to distinguish MPs particles by visual inspection. The particles must not be broken off; they must be clearly and uniformly colored and free of cells and organic structures (Hidalgo–Ruz et al., 2012). The microscope was connected to a computer to capture images, and photographs were taken in zigzag mode until every position was captured (Tang et al., 2018). MPs abundance was calculated as items/kg dry sediment weight (items/kg d.w) and MPs were classified by shape (fibrils, fragments, and films), color (black, multicolored, blue, yellow, red, transparent, pink, green and purple) and size (length: 45-100, 100-330, 330-500, 500-1000, 1000-2000, 2000-5000 μm). In this study, a miniature Fourier transforms infrared spectrometer (Frontier, PerkinElmer, Waltham, MA, USA) was used to identify common suspicious types of MPs. Suspicious MPs identified by visual inspection were randomly selected for verification. The obtained spectra were verified by comparison with the spectrogram database on the instrument. MPs were identified only if the match rate was higher than 70% (Hidalgo–Ruz et al., 2012).

Samples were collected according to the latest quality assurance and quality control standards with strict control measures (Koelmans et al., 2019; Adomat and Grischek, 2020). Collected sediments were preserved in metal samplers and containers to avoid contamination. Non-textile laboratory overalls, masks, and nitrile gloves were worn throughout sample collection, extraction, and identification to avoid the use of plastic-containing laboratory equipment (Masura et al., 2015). All laboratory equipment was rinsed three times with Milli-Q water before and after use (Prata et al., 2020). During the experiments, all solutions, including FeSO₄ and Milli-Q water, were filtered through a 45 µm filter before use. Filtered Milli-Q was used as a blank in the laboratory and processed using the same procedure as the samples (Zhu et al., 2021). The final data were corrected for the average abundance of MPs in the corresponding blank group. These procedures were used to prevent interference from external MPs.

The diversity index D’ (MPs) was calculated according to Equation 1 to estimate the complexity of MP types and sources in mangrove and non-mangrove sediments in ZJB. In total, three types of D’ (MPs), namely size D’ (MPs), color D’ (MPs), and shape D’ (MPs), were calculated based on their shape, color, and size characteristics, respectively (Wang et al., 2019; Huang et al., 2020b; Huang et al., 2021).

Where S is the number of MPs categories, N_i is the number of MPs categorized into the i^th type, and N is the total number of MPs in the sample.

The enrichment index (EI) is the ratio of mangrove MPs abundance to non-mangrove MPs abundance. EI is calculated using Equation 2 (Huang et al., 2021). EI > 1 indicates that MPs are enriched in mangroves and EI< 1 indicates that MPs are enriched in non-mangroves.

Where A_m is the MPs abundance in mangroves and A_n is the MPs abundance in non-mangroves.

Total organic carbon (TOC) calculated by regression line of Equation 3 burning loss rate and soil organic carbon mass fraction.

where x is the burn-off rate and y is the soil organic carbon mass fraction.

Station locations were mapped using ArcGIS 10.2 (Esri Corporation, New York, USA). Data were analyzed using Microsoft Excel 2019 and plotted and analyzed using Origin 2022, including MPs characteristics (abundance, shape, color, size, and diversity) of mangrove and non-mangrove sediments. Confidence intervals were set at 95% for all tests. One-way ANOVA was performed using IBM SPSS Statistics 26 software. MPs shape, color, size, and diversity were analyzed using a multi-way ANOVA (two factors of station location and whether it was mangrove). Values were considered statistically significant when P < 0.05; P < 0.01 indicated highly significant; and P > 0.05 indicated no significance.

Figure 2 displays the variation in MPs abundance in sediments from both mangrove and non-mangrove areas in ZJB region. MPs were detected in all sediment samples collected from the five stations in ZJB. The abundance of MPs was significantly higher in all mangrove sediments than in non-mangrove sediments (P < 0.05). The mean abundance of MPs in non-mangrove sediment samples was 263.67 ± 85.25 items/kg, compared to 618.17 ± 71.75 items/kg in mangrove sediment samples. The non-mangrove sediment sample from NLR (S1) had the highest abundance of MPs (589.17 items/kg), while the lowest abundance (76.67 items/kg) was found in Sino-Australian Garden (S3). Among the mangrove sediment samples, the highest abundance of MPs (891.67 items/kg) was found at Sino-Australian Garden (S3), while the lowest abundance (351.67 items/kg) was found at Potou Primary School (S5). The sediment samples showed the highest and lowest abundances of MPs at S3, which was the most variable of the five stations.

The study analysed the sample sizes of MPs, which were divided into six categories: 0-100 µm, 100-330 µm, 330-500 µm, 500-1000 µm, 1000-2000 µm, and 2000-5000 µm (Wang et al., 2022; Zhang et al., 2022). The size of the MPs did not differ significantly between station (P > 0.05). As shown in Figure 3A , 100-330 µm was the most abundant size category across all five stations, accounting for an average of 37.30%. The total number of MPs < 500 µm at the five stations was 63.36% and 55.31% at non-mangrove and mangrove stations, respectively.

Based on the survey results of the survey, a total of 12 colors of MPs were identified, as shown in Figure 3B . MPs colors did not differ significantly between sampling areas (P > 0.05). Among all MPs in the non-mangrove sediment samples, green was the predominant color (22.28%), followed by black (17.11%), transparent (16.67%), blue (15.05%), and multicolor (14.91%). Other colors accounted for less than 10%. In this study, multicolor (27.51%) was the most abundant in the mangrove sediment samples, followed by transparent and blue MPs, accounting for 15.78% and 9.67%, respectively. Other colors accounted for less than 10%. Orange and brown were the least common colors found in the study in both mangrove and non-mangrove sediment samples.

The relative abundance of different shapes of MPs was shown in Figure 3C . No significant differences were found in the shapes of the MPs detected from the different areas (P > 0.05). All samples contained four distinct shapes: fragments, films, fibers, and sponges (foams). The majority of identified MPs were fragments, accounting for 49.48% and 69.98% of all non-mangrove and mangrove samples, respectively. Fibers were the second most common shape across all stations, accounting for 37.25% and 12.07% of MPs, respectively. Overall, fragments were more prevalent in the samples from this study.

Figure 4 shows typical MPs characteristics and compositions of MPs captured in the sediments by micro-Fourier Transform infrared (FTIR) spectroscopy. Major polymers were found in selected samples of black fragment (A), red fiber (B), transparent film (C), and white fragment (D); with major types including polypropylene (A), polyethylene (B), polypropylene (C), and docosanol (D).

The diversity of MPs size, color, and shape indices, i.e. size D’ (MPs), color D’ (MPs), and shape D’ (MPs), were calculated separately according to equation (1). The results showed the abundance of sediment MPs in the samples ( Figure 5 ). There were no significant differences in the diversity of D’ (MPs), D’ (MPs), and D’ (MPs) in different areas (P > 0.05). The results of this study showed that the diversity of color, size, and shape of MPs in mangroves was changing and mostly showed a decreasing trend. The diversity of color, size, and shape of MPs in non-mangrove samples was 0.77 ± 0.02, 0.72 ± 0.03, and 0.50 ± 0.07, respectively. The diversity of color, size, and shape of mangrove MPs was 0.77 ± 0.06, 0.75 ± 0.02, and 0.45 ± 0.05, respectively. The enrichment index of the mangroves was calculated according to equation (2) and the results showed that for all the mangroves studied the EI values were greater than 1 and the enrichment index ranged from 1 to 12 ( Figure 6 ).

The POC for both mangroves and non-mangroves stations are shown in Figure 7 . It can be seen that particulate organic carbon was higher in mangrove areas than that in non-mangrove stations, with the highest and lowest POC occurring at the same sampling point. In the mangrove sediment samples, the mean POC was 13.12 g/cm³ with the highest POC in NLR (S1) (16.66 g/cm³) and the lowest in LTR (S2) (9.76 g/cm³), and in the non-mangrove sediment, the mean POC was 11.01 g/cm³ with the highest in NLR (S1) (13.72 g/cm³) and the lowest POC in LTR (S2) (9.16 g/cm³). There was no significant difference in POC between mangrove and non-mangrove areas (P > 0.05).

MPs abundance varies between sampling stations in different regions. Human activities have been found to influence MPs abundance, such as agricultural activities, aquaculture (Hinojosa and Thiel, 2009), wastewater discharge (Gies et al., 2018), tourism activities, residential life, tidal action (Zhang et al., 2022), and seasonal variations (Wang et al., 2022). A comparison of MPs abundance in dry mangrove sediments from different regions is shown in Table 2 . The MPs abundance in ZJB is intermediate compared to foreign contamination, including Muara Angke Wildlife Reserve of Indonesia (28.09 ± 10.28 items/kg), Northern Persian Gulf of (19.5 to 34.5 items/kg), Kuala Muda of Malaysia (430 ± 7.234 items/kg), Penaga of Malaysia (940 ± 15.773 items/kg), Seberang Perai of Malaysia (4000 ± 29.174 items/kg) (Naji et al., 2019; Cordova et al., 2021; Tan and Mohd Zanuri, 2023). Compared with mangrove sediments from other regions of China, it was much higher than that of the North Yellow Sea (37.1 ± 42.7 items/kg), Beibu Bay (26.67 ± 9.43 to 239.94 ± 37.80 items/kg), inside the mangrove of Qinzhou Bay (42.9 ± 26.8 items/kg) (Zhu et al., 2018; Li et al., 2018a; Zhang et al., 2022), lower than outside the mangrove of Qinzhou Bay (2174.5 ± 2206.8 items/kg), Pearl River Estuary, South China (851 ± 177 items/kg), the fringe in Futian mangrove in Shenzhen Bay (2835 ± 713 items/kg) (Li et al., 2018a; Zuo et al., 2020; Duan et al., 2021). Each study may have used a different MPs extraction method and different filters will result in subtle differences. In this study, the abundance of MPs in sediments from mangrove areas in ZJB was 1-11 times higher than in non-mangrove areas, a result that is consistent with other studies, suggesting that mangroves trap some of the sediment and do not allow MPs to fully flow into the ocean (Liu et al., 2022).

The shape, color, size, and other characteristics of MPs are key factors in determining the source of MPs. Comparing the five stations, the abundance results showed that the Nanliu River (S1) may be the river with the highest amount of MPs due to the presence of many industrial plants in its vicinity, and the effluents from the plants contained large amounts of MPs, which were discharged into the nearby rivers and thus flowed into the ocean (Zhang et al., 2019). The large difference in MPs between mangroves and non-mangroves in Sino-Australian Garden (S3) may be due to the large distance between the locations of the two sets of samples collected, coupled with the fact that this is a tourist attraction where plastic waste is not properly disposed of and exposed to the environment, thus decomposing into MPs through various effects, with the intermediate anthropogenic factors dominating the results. The MPs in Suixi River (S4) may be due to the high river discharge and fast flow rate, resulting in the input of MPs from the river (Wang et al., 2022). Both the Lvtang River (S2) and the Potou Primary School (S5) were due to the input of MPs from nearby residential life into the river. 80% of the MPs were smaller than 1000 μm in size. All mangrove and non-mangrove samples contained more fragments than other shapes, which could be fragments of MPs left behind by fishing gear during fishing (Li and Liu, 2018). Transparent MPs were the most frequently detected in both mangroves and non-mangroves combined, most likely from transparent plastic products from the daily lives of the surrounding population (Liu et al., 2022).

According to the regression line analysis of MPs abundance and POC, MPs abundance was not significantly correlated with POC (P > 0.05) ( Figure 8 ). It was possible that MPs abundance was not correlated with POC due to the high variability in estuarine flow dynamics, which prevented MPs from aggregating easily in the sediment. It was also possible that the sample data collected was too small to support a relationship between MPs abundance and POC, which needs to be investigated further. In addition, the relationship between MPs and POC could be explored if more samples could be collected from different stations. Galgani et al. (2019) found that plastics have the potential to alter carbon sequestration and carbon turnover, with MPs increasing the production of organic carbon and its aggregation into gelatinous particles, and an increase in gelatinous organics may affect marine bio-pumping and transport of MPs in the ocean, which can affect the ocean nutrient cycling and the global ocean productivity. Blue carbon stocks were not only provided by internal carbon sources, but also by external sources such as phytoplankton, algae, seagrass meadows, mangrove debris, and terrestrial sources (Röhr et al., 2016; Huang et al., 2021). Many studies have concluded that MPs abundance in sediments has a high positive correlation with organic carbon (Chen and Lee, 2021). However, it has also been suggested that MPs abundance was not necessarily positively correlated with sediment organic carbon content and that regional and anthropogenic influences need to be considered (Zhou et al., 2023), and our results were consistent with this. Blue carbon ecosystems, where the presence of submerged vegetation (e.g., mangroves) reduces water velocity through interception to store carbon, may also lead to the accumulation of MPs particles, which can adversely affect vegetation growth and thus its ability to sequester carbon (Huang et al., 2021). In addition, environmental MPs can enter plant tissues and adversely affect photosynthesis and metabolism (Dad et al., 2023), which indirectly affects the global carbon cycle by decreasing the efficiency of plants in sequestering and utilising atmospheric carbon dioxide, contributing to the global greenhouse effect (Li et al., 2024). Based on the experimental results, it indicated that MPs pollution had not yet had a significant impact on the blue carbon system. Both the abundance of MPs and the content of POC were higher in mangrove areas than in non-mangrove areas, and it can be predicted that MPs were much higher in mangrove areas than in non-mangrove areas. This also suggested that mangroves could trap MPs.

MPs are monitored and analyzed using techniques such as microscopy, Fourier Transform Infrared (FTIR) and Raman spectroscopy (Raman), and thermal analysis (Shim et al., 2017). MPs can be monitored in a variety of environments, including water (Li et al., 2018b), air (Gasperi et al., 2018), soil (Jacques and Prosser, 2021), sediment (Uddin et al., 2021), and marine fish (Bråte et al., 2016). MPs can be monitored from the interior and edges of mangroves (Duan et al., 2021), and this study was conducted to monitor MPs in estuarine mangrove and non-mangrove sediments, and the comparative results showed more MPs in mangroves than in non-mangroves, indicating that mangroves can trap MPs, and the trapping effect of MPs was greatly influenced by the adjacent land-based sources induced by human activities. Although wastewater treatment plants can remove more than 90% of MPs (Iyare et al., 2020), discharged MPs can still cause significant environmental pollution, with nearly 48% of wastewater discharged globally without treatment (Jones et al., 2021). This suggests that untreated wastewater is likely to be a major source of MPs in areas where wastewater treatment facilities are limited or non-existen (Rico et al., 2023). Therefore, MPs can be controlled from the perspective of wastewater treatment. Governments should set standards for wastewater treatment, companies should strictly enforce them, and people should raise environmental awareness to reduce the use of plastic products. There are limitations to this study in that the sample size was not large enough, and future studies should focus on organic-rich sediments to further explore the relationship between MPs and blue carbon storage.