Characteristics and distribution of microplastics in shoreline sediments of the Yangtze River, main tributaries and lakes in China—From upper reaches to the estuary

Microplastics (MPs) pervade the environment and increasingly threaten both natural ecosystems and human health. In this study, we investigated MP particle concentrations in sediment samples collected from 54 sites along the banks of the Yangtze River and its major tributaries and on lakeshores. The main polymer types found in the samples were polypropylene (PP), polystyrene (PS) and polyethylene (PE). MP particle abundance in the various types of locations was 35–51,968 particles/kg dry weight (d.w.) on the banks of the main river, 52–1463 particles/kg (d.w.) on the banks of tributaries and 2574–23,685 particles/kg (d.w.) on lakeshores. Correlation between MP abundance and mean annual runoff of each upstream tributary was significant, which suggests that increased runoff brings more microplastic waste to streambank sediments. The most common shape of MP particles in all upstream samples was flake, and in downstream samples it was foam. Small microplastic particles (< 0.50 mm) were predominant at all sites in this study, and the minimum particle size in samples from the Yangtze river banks was 0.065 mm. Average abundance of MP particles on the shores of the source lake was 9069 particles/kg around the inlet but only 866 particles/kg around the outlet; the difference was due to interception associated with sedimentation and precipitation in the lake. Our study represents the large-scale study of MPs contamination in sediment along the Yangtze River and provides important data regarding the accumulation and distribution of MPs in shoreline sediments of the upper, middle and lower reaches of the Yangtze River, main tributaries and lakes in China.


Introduction
Microplastics (MPs) are plastics < 5 mm in size that originate from primary and secondary MPs (Endo et al. 2005;Ryan et al. 2009). Since the concept of MPs was first proposed by Thompson et al. (2004), MPs have received considerable academic attention in recent years due to their persistence and questionable bioavailability in environments (do Sul and Costa 2014;Rachman 2018). MPs are found not only in surface water, but also in deep water, sediment, soil and organisms (Bergmann et al. 2017;Chae and An 2018). In general, small MPs cause the most harm to the natural environment because they are easily available to higher species and are concentrated as they move up the biological food chain (Canesi et al. 2015). MPs can also become vectors of toxic and harmful substances that can cause food or water pollution (Zhang et al. 2018a, b).
With respect to the presence of MPs in water, the importance of marine microplastic pollution research is self-evident (Hidalgo-Ruz et al. 2012;Woodall Lucy et al. 2014;Desforges et al. 2014;do Sul and Costa 2014;Nguyen et al. 2019). At the same time, the study of MPs in freshwater areas is of great significance Wang et al. 2018;Yin et al. 2020;Sekudewicz et al. 2021). Numerous studies have shown that most MPs in the ocean originate on land, and it has been estimated that globally 4.8-12.7 Mt plastic debris from inland water systems and terrestrial ecosystems has been transported to the ocean by riverine transport in 2010 (Jambeck et al. 2015). However, that 262.3-270.2 Mt plastic wastes remain the terrestrial environment, which is widely distributed in river headwaters, basins and estuaries, particularly nearby cities are hotspots of MP pollution Hu et al. 2018;Jang et al. 2020).
As one of the largest rivers in the world, the Yangtze River is considered to be the largest plastic-export river to the ocean in all existing modeling studies (Lebreton et al. 2017;Schmidt et al. 2017) due to its high population density of its catchment and high consumption of plastics in China. However, the current research on the Yangtze River was mainly focused on the middle and lower reaches and major lakes, while relevant research has not been common in the source and main tributaries of the Yangtze River. Moreover, there have been few studies on characteristics and distribution of MPs in shoreline sediments of the Yangtze River, main tributaries and lakes in China. Therefore, systematic research on MPs in shoreline sediments of the Yangtze River, main tributaries and lakes can not only fill the data gap of MPs pollution in the whole Yangtze river valley area, but also provide reference for the better understanding of MPs distribution regularity in the upper and lower reaches of its tributaries, major urban basins and the import and export of main lakes.
The aims of this study are: (1) to illustrate the characteristics, abundance and spatial distribution of MPs in the banks of the upper, middle and lower reaches of the Yangtze River, main tributaries and the shores of lakes of China and (2) to explore the potential sources, influencing factors and their relationship to the distribution and retention of MPs in the Yangtze river valley area shoreline sediments. Results of this study can provide important data on the processes influencing microplastic distribution in shoreline sediments of the Yangtze River, main tributaries and lakes in China. At the same time, it also provides certain data support for local authorities in China to formulate microplastic pollution prevention measures and corresponding laws and regulations.

Study area
Yangtze River, the third longest river in the word, flows through 11 provinces, regions and cities, from glaciers on Qinghai-Tibet Plateau into the East China Sea at Shanghai in China. The Jinsha River is the upper reaches of the Yangtze River, from Qinghai, Tibet into Sichuan province, and the Yalong River is the largest tributary of the Jinsha River. Tuo River, Jialing River and Wu river are also important tributaries of the upper reaches of the Yangtze River, located in southwest China. The Han River, the largest tributary of the Yangtze River, flows through Shaanxi and Hubei provinces and joins in Wuhan area. These tributaries are extremely important hydropower resources in China because of their large river fall and water flow. By 2013, there were 80 large reservoirs and 381 medium-sized reservoirs in the upper reaches of the Yangtze River (Yao et al. 2016). There are about 760 freshwater lakes mainly distributed in the plain of the middle and lower reaches of the Yangtze River with a total area of 17 093 km 2 , including Dongting Lake, Poyang Lake, Chao Lake and Tai Lake and so on. These lakes have accelerated the development of local agriculture, industry, shipping, aquaculture and tourism (Dong and Wang 2013). Samples were collected from Yushu, Qinghai province, to the river estuary at Shanghai, including major urban basins in the upper, middle and lower reaches of the Yangtze River, the upper and lower reaches of tributaries, and the import and export areas of lakes. The sampling site locations are shown in Fig. 1, and coordinates are shown in Table S1. This, to the best of knowledge, is the first work to characterize the spatial distribution of MPs, along with influencing factors, in shoreline sediments along the Yangtze River on a large scale.

Sampling
A total of 162 sediment samples were taken from 54 sites along the Yangtze River banks in major urban basins, along the upper and lower banks of main tributaries and at the entrances and exits shores of major lakes, during May and June 2018. Sediment samples from each sample site were collected using a multi-quadrat mixture method. The samples of each quadrat was collected in a 0.1 × 0.1 m 2 quadrant to a depth of 0.2 m using a narrow stainless shovel (Murray and Cowie 2011). The detailed sampling method is as follows: At each site, three quadrats are randomly selected at 20-m intervals along the S-shaped route, and then, the samples of the three quadrats were uniformly mixed and placed in a fabric bag; approximately 4-5 kg sediment specimens were collected. The samples were sealed and stored in the laboratory at room temperature until required for analysis. Wet sediment samples were placed on the clean stainless steel plate covered with aluminum film and air-dried in a clean laboratory-ventilated kitchen to prevent contamination of airborne MPs in the laboratory. MP particle abundance was determined based on the dry weight of the sediment sample.

Sample and data analysis
Each sample was homogenized in the laboratory without cross-contamination. About 500 g of each sample was sieved in a stainless steel sieve (300 mesh, Φ = 0.050 mm) and leached using pressurized water until the leaching water ran clear (Lin et al. 2021). Sieve residue was collected in a volumetric flask that had been cleaned using ultrapure water and then, for flotation of microplastic particles, mixed and shaken with NaCl-saturated solution (1.12 g/ cm 3 ) and ZnCl 2 solution (1.6 g/cm 3 ) three times successively. The supernatant of each flotation was decanted into a 500-mL glass beaker and vacuum-filtered with 10-20 μm quantitative paper (Φ = 9.0 cm, Xinxing, China). The filter membrane was washed with deionized filtered water, and the residue was carefully transferred to a clean Petri dish using a washed metallic needle before being covered and dried in a drying oven at 50-60 ºC for > 12 h. The dried MPs samples were treated with 30% H 2 O 2 (50 ml) at room temperature in the dark for 48 h in covered glass conical flask (250 ml) to digest the biological and abiological materials and then passed through the 0.45-μm glass microfiber papers (GF /F, 47mmΦ, Whatman), using a vacuum pump. After that, the filter paper was washed with deionized filtered water again and transferred into clean glass petri dish, dried and examined carefully under magnifying glass (PDOK 10X, Germany) and microscope (Olympas, IX81). Particles > 0.15 mm that were visually identified as or suspected to be plastics were transferred from the residues onto clean, black, smooth cardboard and classified according to shape (Nor and Obbard 2014;Zhou et al. 2018). All particles transferred to the cardboard were photographed using a digital camera (Nikon D3200, Japan), and particle counting and size measurement were taken using Nano Measurer 1.2 software (Zhou et al. 2018). MPs were classified the size ranges into < 0.5 mm, 0.5-1.0 mm, 1.0-2.0 mm, 2.0-3.0 mm, 3.0-4.0 mm, and 4.0-5.0 mm.
The MPs extracted from 500 g of dry sediment were divided into 10 parts on average, and two parts were selected to identify the chemical components by attenuated total reflection with Fourier transform infrared spectroscopy (ATR-FTIR). ATR-FTIR spectra were recorded on a Bruker Tensor 27 FTIR spectroscope (Bruker Optics, GmbH, Germany) with Pike Miracle ATR accessories (Pike Technologies, Wisconsin, USA). A set of representative microplastic particles (> 0.15 mm) from each site was selected to identify the chemical component by ATR-FTIR (He et al. 2021). In this research, there were two rules for the selection of microplastic particles: (i) at each sampling site, > 20 particles were selected and (ii) for each type (shape and size) of MPs, > 20 particles were selected . The European Commission's Marine Strategy Framework Directive recommends that at least 10% of recorded fragments and filaments should be analyzed by FTIR (Galgani et al. 2013). In our ATR-FTIR analysis, the proportion of MPs detected accounted for more than 20% of the total microplastics, higher than the 10% regulated by the European Commission, so it was well representative. Some of the MPs might be indiscernible based on their morphologies and so two or three representative MPs of these types were selected for polymer identification (zhou et al. 2018). ArcGIS10.2 (Environmental Systems Research Institute Inc, Redlands, CA, USA) was used to map the sampling sites of MP particles. WPS Office 2019 and OriginPro 8.5 (Origin-Lab Corporation, Northampton, MA) were used for data analysis. Statistical differences were analyzed by one-way analysis of variance (ANOVA) using the SPSS 17.0 software package. All results concerning MP particle abundances are presented as the number of microplastic particles per dry mass of sediment (particles/kg).

Quality assurance and quality control
In order to guarantee the accuracy of our data, a series of quality assurance and quality control (QA/QC) measures were taken during the process from field sampling to laboratory analysis (Brander et al. 2020). During field sampling, all tools were carefully rinsed three times with ultrapure water at each individual step to avoid pollution. The samples were kept in closed spaces to reduce pollution from airborne MPs. Field blank tests were conducted in three randomly selected sampling sites. On each site, 100 g of quartz sand was washed by pressurized distilled water and passed through a 50-μm stainless steel sieve (300 mesh). The duplicates were set in triple. Subsequent treatment for blank samples was identical with that for shoreline sediments as above mentioned. Finally, only 3.15 ± 1.24 plastic particles were detected under the microscope in the blank samples. The results showed that potential microplastic contamination was minimal and can be negligible.
To avoid potential background contamination during the analysis process in the laboratory, the following preventive measures described in the literature were taken in this study. MP particles were separated and enumerated in different clean rooms. Cotton laboratory coats were worn at all times during analysis. All materials and vessels were covered with aluminum foil after each individual step. All liquid used in the experiment was filtered through the 0.45-μm filter (GF/F, 47mmΦ, Whatman). During the processing of MP in the laboratory, procedural blank samples (3 ultrapure water samples and 3 NaCl solution samples) were included. Potential airborne particles were examined by sucking the air in the workplace through the 0.45-μm filter paper for 2 h under vacuum filtration, and no plastic particles were found under the microscope. Each test was set in triple. The results of blank samples indicated that the contamination from transportation to processing was negligible.
The MP recovery of the extraction procedure was also ascertained by 100 of MP particles of dimensions (0.1-5.0 mm) into clean artificial sediment (quartz sand; 0.25 mm; 500 g) and processing under the same conditions that were described above for environmental samples. Same procedure was repeated for three times. The recovery rate was 93.5 ± 4.3%.

Shape categories, polymer type and distribution of MPs
According to their shapes and features, the plastic/microplastic particle were categorized into pellets, foams, fragments, flakes, films, fibers and sponges, as shown in Fig. 2. In general, pellet was hard, regular, disc-shaped, oval or cylindrical plastic particles. Foam was white expanded polystyrene particles. Film was thin, soft and light-transmitting particles of plastic sheet. Fiber was long, curly remnants of dust screen or fishing nets. Fragment was hard, thick pieces of plastic with irregular shape. Flake was usually opaque, regular flat particles of various plastic woven bags. Sponge was particles that were yellow, porous and irregular honeycomb. The composition and classification of sediment samples were determined by attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectrometry using the detected spectra, as shown in Fig. S1.
Flake particles decreased from 46.3% of MPs in the upper reaches of the Yangtze River bank sediments to 10.2% in the lower reaches. In the middle reaches, the proportion of flake particles was 68.0%; the greatest abundance was in the Wuhan section, 88.1%. In contrast, foam particles were 4.9% in the upper reaches of the Yangtze and 59% in the lower reaches, a tenfold increase, as shown in Fig. 3(a). In the tributaries of the upper reaches of the Yangtze River, particles were primarily flake (maximum 76.1%) and film (maximum 46.0%), and the average proportion of foam particles was < 10%. No spherical particles were found in the tributary sediments, as shown in Fig. 3(b). Particles representative of all seven shape categories were found in the lakeshore sediments: the average proportion of foam particles was 41.7%, with a maximum of 64.4% for Poyang Lake; the average proportion of flake was 20.05% and of film, 13.8%, as shown in Fig. 3(c).

Abundance and size of MPs in the Yangtze River basin
The abundance of MP particles in the Yangtze River banks is shown in Fig. 4(a). The abundance of MP particles in the beach sediments of the main urban basins along the Yangtze River varied greatly. The abundance of MP particles in riverbank sediments from Chongqing to Yueyang was relatively small, in the range 35-322 particles/ kg. However, from Wuhan to Shanghai, the abundance of MP particles was relatively high, mainly in the range 278-51 968 particles/kg. The greatest abundance of MP  particles in sediment samples was found in the Wuhan region, followed by the Jiujiang, Shanghai and Nanjing regions. The abundance of MP particles in the Yangtze estuary was 992 particles/kg. Abundance in the upstream tributaries was in the range 254-1463 particles/kg, as shown in Fig. 4(b); the highest abundance was found in the Ming River bank. The abundance of MP particles in the shores of Dongting Lake (2576 particles/kg), Poyang Lake (8017 particles/kg), Chao Lake (5460 particles/kg) and Tai Lake (23 685 particles/kg) was significantly greater than in the banks of the tributaries, as shown in Fig. 4(c). Zhao et al. (2014) and He et al. (2021) reported that MPs with particle size < 0.5 mm were the most abundant MPs in the surface waters of the Yangtze River, which was attributed to long-term degradation of small plastic debris. The size distributions of MPs in sediment determined by our experiments are shown in Fig. 5. The size range of MPs detected in the sediments of the main stream was 0.065 ± 0.013-4.02 ± 0.452 mm, in tributaries 0.071 ± 0.008-3.91 ± 0.883 mm, and in lakes 0.073 ± 0.016-4.10 ± 0.593 mm (Table S2). MPs < 0.5 mm were predominant in the sediments of the Yangtze River banks (average 66.7%), tributary riverbanks (average 64.1%) and lakeshores (average 65.9%); the greatest proportions were 88.1% in the Wuhan region, 70% in Tuo River banks and 79.3% in Chao Lakeshores. The general trend was that larger MP particles (> 2 mm) were less abundant in the sediments of the Yangtze River banks (in the range 0.0-12.85%), tributary riverbanks (in the range 3.9-7.5%) and lakeshores (in the range 5.5-24.56%). We found that sample sites with high abundance of flake (PP) and foam (PS) particles had higher proportions of small particles.

MPs contamination in tributary riverbanks
The distributions of MP particles showed different trends upstream and downstream in each tributary (Table S3 and Fig. S2). Flake particles were predominant in upstream samples, accounting for 73% on average. The greatest proportion of flake particles was 98.8% in the Ming River samples. The proportion of fiber particles was 74.3% in upstream samples of the Tuo River. No pellet or sponge particles were found in the upstream samples. Downstream in the tributaries, film (average 39.2%), flake (average 20.1%), fragment (average 13.5%) and foam (average 14.4%) were the most common particle shapes found in the sediment samples. Fiber particles were found in the Jialing (41.8%) and Wu River (43.3%) samples. Sponge particles were found only in the Jialing samples (3.6%), and no pellet particles were found in any tributary samples. Abundance of MP particles in samples from the upper reaches of the tributaries was in the range 192-1004 particles/kg (average 418 particles/kg), and samples from the lower reaches were in the range 30-580 particles/kg (average 252 particles/kg) (Fig. 6a).

Microplastic contamination characteristics of lakeshores
Flake particles accounted for 32.6% on average, followed by fragment (24.2%), foam and film accounted for 21.5% and 19.2%, respectively, in the Dongting Lake samples. Foam, flake and fiber particles accounted, respectively, for 27.3%, 36.4% and 26.8% of the Chao Lake samples. Foam particles were predominant in samples from both Poyang Lake (58.4%) and Tai Lake (70.4%). Foam particles were widespread; we found them in every lake sample, with abundance in the range 21.5%-70.4%. At the outlet of Dongting Lake, film and flake particles accounted for 90% of particles found in the samples. Foam particles accounted for 88.5% of the particles found in the Poyang Lake outlet samples. Foam, fragment, film, fiber and flake particles were predominant in both the Chao (81.7%) and Tai Lake outlet samples (82.9%). Pellet and sponge particles had an abundance of < 10% in the inlet and outlet samples of these two lakes (Table S4 and Fig. S3). The abundance of MP particles differed significantly between inlet and outlet samples for all lakes, as shown in Fig. 6(b). The abundance of MP particles in lake inlet samples was in the range 2546-21,112 particles/

Correlates of MPs in the Yangtze River basin
We found significant correlations between particle abundance in tributary samples and water runoff (Fig. 7a and Table S5). The abundance of flake particles along the Yangtze River banks was positively correlated with the total MP abundance in the main tributary ( Fig. 7b and Table S6). There were significant correlations between the abundance of foam particles and total MP particle abundance in the main sampling areas of the Yangtze River ( Fig. 7c and Table S7).

Source and distribution of MPs
There may be a close relationship between human activities along the Yangtze River and the different types of MP particles found in our sampling (Wang et al. 2017;Yang et al. 2021). Different factors contribute to the spatial differences observed in MP contaminants along the river. We frequently observed discarded foam floats and containers, plastic drums and bottles, fishing nets, woven plastic bags and food wrappers on the banks of the Yangtze River and its main tributaries and on the lakeshores during field sampling, all of which are potential sources of MP particles. One-way analysis of variance (ANOVA) with Dunnett T3 post hoc test was performed to determine differences in particle number among different treatment groups (a, b, n = 3). The letters above the bars indicate significant differences (p < 0.05). Different capital letters represent significant differences between upstream and downstream of the same tributary / inlet and outlet of the same lake, and different lowercase letters represent significant differences between different tributaries / lakes at the same site (c) abundance of foam MPs and total MPs in the main sampling areas of the Yangtze River Flake particles predominated in the upstream Yangtze River samples. These have originated from gradual damage to discarded plastic woven bags. Polystyrene foam particles were found mainly in the Yangtze River bank samples; these originated primarily from aquatic product packing boxes and discarded foam floats, as reported by Free et al. (2014) and Lee et al. (2015). Fiber particles were widely distributed in the Yangtze River bank samples, mainly due to the large use of wind fences in the economically less developed areas of the upper reaches of the Yangtze River, and to the fishery that has developed in the lower reaches of the river. Film particles originated mainly from the degradation of discarded plastic bags and agricultural mulch (Sintim and Flury 2017). The main sources of fragment particles were discarded objects that resulted from human lifestyles and the industrial production of hard plastic products; over the long term, hard plastic products break and produce fragments (e.g., broken plastic tubs, buckets and bottles). Film and fragment particles were widely present in samples throughout the survey area, with a relatively high distribution in more populated areas that further illustrates the close relationship between MPs and human activities. Pellet particles were found in the middle and lower reaches of the Yangtze River bank samples, which we mainly attributed to improperly discarded industrially manufactured primary materials. More pellet particles were found in lakeshore sediments because a lake is an important destination for pellet particle migration in the drainage basin. Sponges are widely used in industry and commerce for shock absorption, packaging and heat insulation, and they increase pollution by PEU MPs (Zhou et al. 2018).

Microplastic contamination along the Yangtze River
MP particle concentrations in different reaches of the Yangtze River were dissimilar because they were influenced by different environmental conditions and had different sources (Feng et al. 2021). Previous studies have suggested that MP concentrations in aquatic environments were mainly driven by population density or economic development (Lahens et al. 2018;Peng et al. 2018;Eriksen et al. 2013;Yonkos et al. 2014;Liu et al. 2019). We found that the abundance of MP particles along the Yangtze River banks was relatively high in major urban watersheds, but there was no significant correlation between abundance and local population size or economic development indicators. We conclude that MP particles in shoal sediments have a complex of sources, including local discarding of plastic products and continuous input from upstream rivers, and that local hydrological conditions maybe have an important effect on plastic deposition. The abundance of MPs in riverbank sediments is also a consequence of long-term accumulation. Over time, different types of plastic products and MPs are weathered by various factors, a process that continuously increases MP particle abundance (Song et al. 2017). In addition, lower density particles of polymer PE, PP and expanded PS that are distributed on the surfaces of rivers and lakes are more likely to be trapped in the riverbank and shoreline sediments because they are influenced by tides and wind, thus increasing the proportion of small MPs in the sediments of riverbanks and lakeshores. The Yangtze River has long been considered to be the largest source of MPs entering the ocean, but previous studies ignored the retention of MPs in shoal sediments, leading to overestimates of the flux of MPs from rivers into the ocean (Siegfried et al. 2017).
In contrast to previous studies of MP accumulation, we found that the abundance of MPs in the sediments of the main Yangtze River basin is at an intermediate level (Ding et al. 2019;Lin et al. 2018;Klein et al. 2015). A single shape category of MPs is common in the upper reaches of the Yangtze River, and the abundance of MP particles is also relatively low, due to less human activity in the region. However, in the middle and lower reaches, as opposed to the upper reaches, the shape category and abundance of MP particles were greater. Among them, the greatest abundance was found around Wuhan, and the particles were mainly flakes. This is related to the long-term use of plastic woven bags in local long-term flood defenses and remediation. There were more foam and fiber MP particles in the lower reaches due to the developed fishery and transportation of aquatic products. The significant positive correlation between the abundance of foam particles and the total abundance of MP particles implies that PS plastic products are the main sources of MPs in the sediment samples of the main sampling areas of the Yangtze River. MP particle abundance was relatively high in the Yangtze estuary, which indicates that urban estuarine river input was an important source of MPs entering the ocean (Rech et al. 2014;Zhao et al. 2015). Tides may also be an important factor in the accumulation of MPs in estuaries (Xiong et al. 2019).

MPsin the tributaries of the upper Yangtze River
In the tributaries of the upper reaches of the Yangtze River, where human activities are less relevant, the abundance of MP particles is nonetheless relatively high. This is related to the types of polymer; Song et al. (2017) reported that polypropylene (PP) and polystyrene (PS) are more prone to fragmentation under natural conditions. The significant positive correlations between the abundance of flake particles and the total abundance of MP particles in tributary riverbank samples indicate that flake particles are predominant.
In the upper reaches of the tributaries, which are undeveloped, woven plastic bags and netted wind fencing discarded during the construction and maintenance of water conservancy, and hydropower facilities produced a high abundance of MP particles. Particle shapes were mainly flake and fiber, and flake particles were predominant. Agriculture and other human activity are more frequent in the lower reaches of each tributary, and both overall abundance and the number of film and fragment categories in our samples increased. However, flake particles decreased significantly in the lower reaches of each tributary. This observation suggests that the dam on the upper reaches of a tributary may have trapped considerable amounts of MPs that originated upstream of the dam (Zhang et al. 2015(Zhang et al. , 2017Xiong et al. 2019;Lin et al. 2021). This suggested that different human activities can affect the types, abundance and spatial distribution of MPs in the upper and lower reaches of tributaries.
Runoff is believed to contribute to microplastic pollution (Gasperi et al. 2014;Dai and Zhang 2013). A significant correlation between MP particle abundance and mean annual runoff was found for each tributary of the upper Yangtze River. This suggests that increased runoff transports more surface microplastic waste into the rivers, and as a result, the abundance of MP particles remaining in riverbank sediments may be increased by waves or wind. These particles will, over time, flow into the upper reaches of the Yangtze River and continue to be transported downstream, possibly into the oceanic system, thus becoming an important source of marine MPs.

Characteristics of MPs contamination in lakeshores
The sediment MP particle concentrations found in lakeshore samples were high in comparison with some other studies, such as of the Caofeidian reclamation area (Zhou et al. 2016), Halifax Harbor in Nova Scotia, Canada (Mathalon and Hill 2014), the Poyang Lake system (Jian et al. 2020), Dongting Lake  and Tai Lake (Su et al. 2016).
Particles in all seven shape categories were found in all lakeshore sediment samples. Abundance of MP particles in lakeshore samples was significantly greater than in riverbank sediment samples because particles in rivers are mainly transported downstream, whereas particles in lakes are mainly deposited and thus remain in place. Abundances in the lake outlet samples were significantly less than those found at inlets. This is because the lake water outflow rate was low; still water provides good conditions for a large number of MP particles to be trapped in lake and lakeshore sediments (Xiong et al. 2019). Our findings imply that previous models may have overestimated the flux of MPs from land to ocean because they did not take into account the retention of MPs in river and lake systems.

Conclusions
Our results show the ubiquity of MPs in sediments from the banks of the Yangtze River and tributaries and lakeshores. We also characterized the accumulation and distribution of MPs in the survey region. We identified seven shape categories across the entire Yangtze River basin and analyzed the main polymer types as polypropylene (PP), polystyrene (PS) and polyethylene (PE). The sources of the different categories of MPs were closely related to human activity along the Yangtze River. The higher abundance of MPs near urban centers and large lakes that are part of the Yangtze River system, and in the river estuary, indicates that both human activity and large bodies of open water affect MPs density in shore and bank sediment. Flake particles were the major constituents in all upstream samples, and foam particles were predominant in downstream samples. The distribution of foam particles displayed an east-west pattern of stepwise descent, but flake and film particles showed the opposite distribution. Small MP particles (< 0.50 mm) were predominant at all sites in this study; the minimum size of particles in the Yangtze River banks was 0.065 mm. There was a significant correlation between MP particle abundance and mean annual runoff of the tributary basin in the upper Yangtze River tributaries. There was an order of magnitude reduction in average abundance of MP particles in lakeshore sediments between lake inlet and outlet due to particle interception associated with sedimentation and precipitation in the lake. Systematic surveys of hydro-sedimentary parameters and better characterization of primary plastics sources are needed to better understand the hydrodynamic drivers of the distribution and persistence of plastic particles in the Yangtze River. The MP particle abundances in sediments can act as a proxy for the spatial extension of human activities that are propagated across the Yangtze River basin, and they are also a fingerprint of pollution sources.