Litter
A total of 11,179 litter items (average = 139.74 ± 20.69 SE items/100 m), weighing 20,028 g (20 kg) (average = 249.96 ± 66.14 SE g/100 m) were collected in the sampling period (summer and autumn at both Woodbridge Island and Derdesteen). Average litter concentrations were significantly higher (U = 1522, p = 0.01) at Woodbridge Island (250.10 ± 33.10 items/100 m) than Derdesteen (29.40 ± 4.31 items/100 m). Plastic was the most abundant litter type found during this study with over 90% and 47% of all litter by count and weight, respectively. Of the plastic litter, foam was the most abundant type by count and weight (41% and 48% respectively). The number and weight of litter found in this study were considerably lower than other studies in Table Bay such as Opie (2021) who recorded 39,602 items weighing 116.6 kg in 2019. It is well documented that estimates of litter accumulation are influenced by sampling methodologies (Eriksson et al., 2013; Ryan et al., 2014; Smith & Markic, 2013). Opie (2021) sampled litter during winter, spring, and summer while the present study only sampled for summer and autumn. In addition, Opie (2021) survey was carried in 2019 pre-Covid19 pandemic while this survey was carried in 2021, during the pandemic.
Total litter abundance differed between summer and autumn for both sites sampled. Litter count was higher in autumn (5,883 items, 52.63%) than summer (5,296 items, 47.37%) (Fig. 3a). Based on total weight, litter was heavier in summer (15,132 g, 75.55%) than in autumn (4,895 g, 24.45%) (Fig. 3b). Opie (2021) also found seasonal variations in litter at Table Bay, where more items were found in winter compared to other seasons. In Chile, a total of 19,886 litter items were collected and seasonal variations between fall (11,894 items representing 59.8%) and winter (7,992 items representing 40.2%) was observed (Barría-Herrera et al., 2021). The present study was carried out during summer and autumn, when beach usage in the summer months (December-February) in Cape Town is more intensive, potentially resulting in greater litter abundance. Due to the COVID-19 pandemic and the enforced restrictions on the movement of people, access to the beach for recreational purposes in South Africa was prohibited in November 2020 to the end of January 2021, implying fewer visits during that period. It is important to note that beach clean-ups resumed as soon as the beach reopened.
The collection of litter for the summer was done in early February 2021, just as access to the beach was allowed. This may explain why fewer items were collected in the summer compared to the autumn season.
This study showed inter-beach differences in litter collected in both summer and autumn. Total litter items by count were higher at Woodbridge Island with 10,004 items (89.49%) compared to Derdesteen 1,175 items, 10.51%) (Fig. 4a). Litter items found in Woodbridge Island were also heavier, with 12,341 g/100 m (61.62%) compared to 7,686 g/100 m (38.38%) in Derdesteen (Fig. 4b). The variation of litter items and weight could be explained by firstly, the difference in land use between Woodbridge Island beach (mainly residential, adjacent to an industrial area and downstream from 2 river mouths) and Derdesteen (nature reserve) (Fig. 1). Chitaka & von Blottnitz (2019) and Lamprecht et al. (2013) identified public access and proximity to metropolitan areas as influential factors in marine litter occurrence. This could explain why Woodbridge Island had more litter than Derdesteen. Secondly, the proximity to population centres may be the cause of litter accumulation, as beach debris is proportional to the number of beachgoers (Barnes et al., 2009).
Although there was no correlation between wind speed and litter count (r = -0.04, p = 0.722) nor wind speed and litter weight (r = 0.111, p = 0.327) during the sampling period, studies have identified wind as a factor influencing debris transport (Faure et al., 2015; Ivar do Sul & Costa, 2007). Studies in South Africa by Swanepoel (1995), Lamprecht et al. (2013) and Opie (2021) reported that wind influences the variability of litter between sites in Cape Town. Milnerton (Woodbridge Island) and Koeberg beaches of Table Bay, in 1995, 2012, and 2019 were surveyed by Swanepoel (1995), Lamprecht et al. (2013) and Opie (2021) respectively, and more litter at Milnerton relative to Koeberg were recorded.
The results of this study also showed an intra-site (wet vs dry zones) variability in litter in Table Bay. More litter items by total count were collected in the dry zone, with 6,212 items (representing 55.57%) compared to 4,967 items (44.43%) in the wet zone (Fig. 5a). The same trend was observed for total litter weight. In the dry zone, 15,740 g (15.7 kg) (78.59%) were recorded when compared to the wet zone (4,287 g (4.3 kg) (21.41%)) (Fig. 5b). The difference between wet and dry zones could be attributed to the displacement and exposure of buried items by wind as well as littering by beach goers. Meakins et al. (2022) also found more litter items above the tide line (dry zone) in the eThekwini municipality (Durban, South Africa). This, like in the current study was attributed to beach goers as well as wind transporting debris from the wet to the dry zone. In Kenya, Okuku et al. (2020b) collected litter at six beaches and reported higher litter count in the dry zone compared to the wet zone. This was attributed to the presence of a few yet heavy items on the wet zone (shoes, fishing nets, clothes) which could easily be deposited when wave energy is reduced while the dry zones had mostly lighter items (plastic fragments, food wrappers, cigarette butts, lollipop sticks and straws) which were mainly windblown Okuku et al. (2020b).
Daily variations in litter were noted during the two sampling periods (summer and autumn) on Woodbridge Island and Derdesteen in Table Bay (Fig. 6). This could be attributed to the effect of wind and wave action on litter accumulation. However, the observed daily variations in litter count and weight were not significant at the sites during the sampling periods (KW p > 0.05). Daily variations in litter on five (5) beaches in Cape Town have been reported previously (Chitaka & von Blottnitz, 2019). Chitaka & von Blottnitz (2019) attributed this to some extent to varying weather patterns such as rainfall and wind. Rainfall events was reported to have increased the flow of litter via stormwater systems at certain sites, while on other sites, wind action was responsible for blowing away accumulated litter (Chitaka & von Blottnitz, 2019). Lamprecht et al. (2013) also observed daily litter variations in Table Bay in 2012. Wind and wave actions may have been responsible for the daily litter variations observed in this study, by blowing/carrying away accumulated litter and uncovering previously hidden items. Other studies also highlighted that the transportation of litter via wind or water may be influenced by weather patterns (Chitaka & von Blottnitz, 2019; Li et al., 2018). Numerous studies have found a link between rainfall and litter loads (Lee et al., 2017; Rech et al., 2014). However, the results of this study cannot support or reject the impact of rainfall on litter loads because no rainfall events occurred during the collection period. But wind could have acted as a vector for litter transportation in Table Bay, with wind speed, coupled with the direction, resulting in a cleaning effect on the beach.
Brand auditing showed that over 90% of litter found in Table Bay were of local origin. This supports the finding by Ryan et al. (2020), who indicated that most marine debris in South Africa come from local sources. This further suggests that most land-based plastics do not disperse far from source areas (Ryan et al., 2020). In Kenya, Okuku et al. (2020b) also found that most litter collected were of local origin, making up to 88% of all collected items. The brand audit of this study also showed that identified items in Table Bay were mainly of snack packaging (78.95%) and the rest of beverages (21.05%), probably littered by beach goers or litter flushing into adjacent coastal waters from the two rivers upstream of Woodbridge Island.
The result of the clean cost index (CCI) categorized both Woodbridge Island (1.67) and Derdesteen (0.19) within the “very clean” category. This could be attributed to the clean-up efforts from the municipality as observed during the survey as well as the clearing effect of wind as identified in other studies (Chitaka & von Blottnitz, 2019).
Items sampled were categorised into eleven litter types (Barnardo et al., 2021), with plastic being the most abundant type by count (92.27%) and weight (43.45%) (Fig. 7). This was expected as most studies on marine litter worldwide have shown that plastic accounts for 50 to 90% of all marine litter (Agamuthu et al., 2019). Lamprecht et al. (2013) surveyed Milnerton and Koeberg beaches in Cape Town in 2012 and found that plastic was the most abundant marine litter with 93.3% by count and 58% by weight. Chitaka & von Blottnitz (2019) similarly found plastic to be the most abundant litter type during a survey of five beaches in Cape Town in 2017. Plastic items accounted for 94.5–98.9% of the total count and 57.0–83.4% by average weight (Chitaka & von Blottnitz, 2019). The types of litter found in this study mostly align with the global characteristics of litter as mentioned in Galgani et al., 2015).
There was a positive correlation between the weight and count of litter items in Table Bay (r = 0.666, p < 0.001). Chitaka & von Blottnitz (2019) also found a positive correlation between litter weight and count in Table Bay (r = 0.95, p < 0.05).
Of all the plastic litter recorded, foam (mainly polystyrene) was the most abundant both by count and weight, 41.74% and 48.32%, respectively (Fig. 8a and b). However, based on the variety of plastics analysed by FTIR, polyethylene (PE) accounted for 69.23% of polymer types, followed by polypropylene (PP) at 14.10%; Cellulose acetate (CA) and polystyrene (PS) were also identified (Fig. 8c). Schwarz et al. (2019) also found polyethylene to be the most dominant polymer type in their study.
Mesolitter
Mesolitter (5–25 mm) were sampled in summer 2021 at Woodbridge Island and Derdesteen, with a total of 1,428 items/m recorded in sediment samples, weighing 12.20 g/m. The average count was 4.46 ± 0.60 items/m, and average weight was 0.04 ± 0.01 g/m. Ryan et al. (2018) reported an average density of mesolitter of 708 items/m, which is higher than what was found in this study, possibly due to the method and area covered; 82 beaches in 1994, 2002 and 2015 were sampled for the study.
Inter-beach differences in mesolitter collection were found, with Woodbridge Island accounting for 1,406 items/m (mean = 11.72 ± 1.37 items/m), representing 98.46% of the mesolitter count, compared to Derdesteen with 22 items/m (mean = 4.40 ± 1.17 items/m), representing 1.54% (Fig. 9a). In terms of weight, almost the totality of weight was found on Woodbridge Island with 12.20 g/m (average = 0.06 ± 0.02 g/m) representing 99.9% of the total weight (Fig. 9b). Similar to litter, the inter-beach variation of mesolitter could be attributed factors such as public access, proximity to population centre, and wind. Woodbridge Island beach is in a residential area, near the Cape Town harbour and Derdesteen beach in a nature reserve (Fig. 1). Tourism and the number of beach goers have been reported to influence the occurrence of mesolitter (Leite et al., 2014; Okuku et al., 2020a; Poeta et al., 2016). A similar pattern was observed in Kenya where mesolitter count was higher on sites near populated areas compared to those near semi populated, and remote ones (Okuku et al., 2020a).
Furthermore, intra-beach variability was observed in this study, where more mesolitter items by count were recorded in the wet zone (4.68 ± 0.94 items/m/day) compared to the dry zone (4.25 ± 0.76 items/m/day) (Fig. 10a). However, no clear variability was observed for the mean weight of mesolitter items, with approximately 0.04 ± 0.02 g/m/day in both the wet and dry zones (Fig. 10b). In addition, there was no significant differences in the mesolitter count (U = 14 021 p = 0.093) and weight (U = 12 727, p = 0.756) between the two zones. The observation in this study is different to that of a similar study, where the backshore (designated as the dry zone in this study) had the highest mesolitter count (Lee et al., 2017). The intra-beach variabilities of mesolitter items could be attributed to factors such as wind, amounts and sources of items entering the marine environment (Ryan et al., 2020) as well as land use and proximity to metropolitan areas (Chitaka & von Blottnitz, 2019).
The composition of mesolitter was mainly plastics by count (99.05%), of which 65.75% were pellets and pellets also accounted for 78.69% of mesolitter weight, with metals accounting for 21.31% of the remaining weight (Fig. 11a and b). A similar trend was observed in South Africa from a survey of 82 beaches in 1994, 2005, and 2015, with plastic count accounting for 99% by count and 95% by weight (Ryan et al., 2018). In this study, pellets accounted for 69.75 % of all meso-plasics by count, followed by foam (mainly polystyrene) (18.49%) and fragments (10.78%) (Fig. 11a). It is likely that given the low density and size of mesolitter, that wind played a significant role in displacing and uncovering previously buried items. The results of this study highlight the ubiquitous nature of plastics of varying sizes in the marine environment and points to the already observe trend of plastic entering the environment (Agamuthu et al., 2019; Borrelle & Law, 2020). However, the assumption that most mesolitter are sourced from local, land-based sources (Lee et al., 2017), could not be adequately supported by our findings, as pellets dominated mesolitter and higher (non-significant) counts were recorded at the wet zone relative to the dry zone. An indication that the recorded pellets were marine sourced, possibly from nurdle spills at Sea/harbour as the case at Durban harbour, where over 2 billion nurdles were spilled in 2017 (Strydom et al., 2020). Wind action could however influence the variability between the zones, by the displacement of litter on the beach according to the wind direction (Okuku et al., 2020b).
Polymer identification showed that polyethylene (PE) was the most common polymer type, accounting for 59.52%, followed by polypropylene (PP) at 27.38%, polystyrene (PS) at 12.50%, and cellulose acetate (CA) at 0.60% (Fig. 11c).
Microplastics
A total of 688 microplastics (MPs) were extracted from water and sediment in summer (57.81% of samples) and autumn (42.19% of samples) of 2021 in Table Bay (Fig. 12a). More MPs were recorded at Woodbridge Island (75.66%) than Derdesteen (24.34%) (Fig. 12b). The higher abundance of MPs in Woodbridge Island reflects the beach’s proximity to the Diep River mouth (within 1 km), ease of access through public transportation, and presence of food establishments, and proximity to residential area. This is not the situation at Derdesteen beach, which is in a nature reserve, far from residential area and river mouth. Sparks & Awe (2022) reported that site specific variations of MPs in coastal sediment was related to potential sources of MPs, such as stormwater pipes. The presence of MPs in sediment is hence linked to proximity to sources of MPs, including rivers, sewage outfalls, and WWTWs (Claessens et al., 2011; Karthik et al., 2018)
Fibre was the most abundant type of MPs collected from water and sediment samples (97.37%) (Fig. 13a).
The result of this study is similar to Nel & Froneman (2015) who sampled along the south-eastern coastline of South Africa (fibres: 90% and blue/black colour: 90%), and Claessens et al. (2011) along the Belgian coast (fibre: 59%). Chen & Chen (2020) also found fibre to be the most prevalent MP type in Taiwan (fibre: 97–99%). Blue (46.46%) and black (19.62%) were the most common colours recorded (Fig. 13b). Variation of MP colours point to the origin of the particles in the environment (Bimali Koongolla et al., 2018; Dhineka et al., 2022; Stolte et al., 2015). It has been reported that differences in collection methods, lab methodologies, and quantitative units make it difficult to establish a direct and meaningful comparison of results between studies (Nel et al., 2017). In the size category, MPs between 0.1 and 0.5 mm were the most abundant, accounting for 50.32% of MPs sampled (Fig. 13c).
The accumulation of the various plastics assessed at the sites were largely influenced by wind and wave actions, beach goer activities and the coastal catchment areas. Plastics were more prevalent at the Woodbridge site relative to the Derdesteen site. The high population density, public transport access and stormwater inputs (from industrial and residential areas) were identified as important plastics inputs, at the Woodbridge site. Hence, recorded plastics originated mainly from land-based sources.