Lagoons as Ocean Gatekeepers? Seasonal Transport and Retention Dynamics of Floating Macro-debris in the Umgeni River, South Africa

Lagoons as Ocean Gatekeepers? Seasonal Transport and Retention Dynamics of Floating Macro-debris in the Umgeni River, South Africa | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Lagoons as Ocean Gatekeepers? Seasonal Transport and Retention Dynamics of Floating Macro-debris in the Umgeni River, South Africa Tadiwanashe Gutsa, Cristina Trois, Thomas Mani This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9052696/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Rivers are major transport routes for land-based ocean plastic pollution. Spatiotemporal, geomorphological and hydrological factors that influence river debris passage are poorly understood. To address such knowledge gaps, we employed GPS drifters (n = 66 + 4 pilot) mimicking macroplastics (> 5 cm) across three different rainfall phases in the Umgeni River, South Africa. We analysed the drifter trajectories to investigate seasonal transport and retention to estimate debris emission rates into the Indian Ocean. Observed drifter trajectories showed modest seasonal differences in mean daily transport and retention counts. Our pilot drifters captured the 1:50–100 year return period flood occurring in April 2022 showing substantial flushing downstream to the Indian Ocean. Mean retention durations showed a notable decrease from the dry phase towards both the peak wet and the wet–dry transition phases, indicating increased debris mobility during the wet season. Drifters were retained frequently in upstream river sections along meanders and vegetated banks, but the downstream estuarine lagoon emerged as the dominant sink with long-term retention limiting drifter exports into the ocean. Our study presents spatiotemporal insights of macroplastic debris transport to inform river plastic transport modelling and effective debris cleanup and mitigation policy frameworks. Earth and environmental sciences/Climate sciences Biological sciences/Ecology Earth and environmental sciences/Ecology Earth and environmental sciences/Environmental sciences Earth and environmental sciences/Hydrology Earth and environmental sciences/Ocean sciences Macroplastic Pollution Rivers Ocean Seasonality Retention zones Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1.0 Introduction Plastic pollution in waterbodies poses threats to human livelihoods, local economies as well as marine and freshwater ecosystems [ 1 ]. An expected tripling of plastic use and waste by 2060 will likely exacerbate the leakage of mismanaged plastic into freshwater systems and oceans [ 2 , 3 ]. Urban River catchments in regions characterized by dense population and inadequate waste management infrastructure are especially prone to plastic pollution [ 4 , 5 ]. These contexts coupled with seasonal weather and river hydrodynamics introduce complexity in understanding the transfer of waterborne macroplastic from rivers to the ocean [ 6 , 7 ]. Rivers play a dual role: on the one hand they are primary transport agents of litter to the sea (0.8–2.7 million metric tonnes/year) [ 8 ] and on the other hand they act as temporary storage reservoirs [ 9 – 11 ]. Recent papers investigate the transport of floating plastic debris through river networks with the interplay of river hydrodynamics [ 12 – 14 ]. But the varying intricate river catchment characteristics and uncertain longitudinal displacement probability across different landscapes limits our current comprehensive understanding of river plastic transport. Recent studies employ Eulerian or Lagragian methods to investigate riverine macroplastic transport processes. Eulerian methods provide valuable insights into localized transport rates and the extent of floating litter in rivers but are limited in capturing broader spatiotemporal aspects including transfer probability and fate [ 7 , 15 ]. These limitations are particularly pronounced when considering the effects of seasonal, hydrological, geomorphological and tidal forces [ 7 , 16 , 17 ]. Lagrangian approaches offer high resolution tracking of floating debris in rivers, capturing key processes such as temporary retention in floodplains or riverbanks, upstream-to-downstream debris movement and debris transfer dynamics between aquatic domains e.g. river mouth and sea. Recently some pivotal steps towards unravelling plastic pollution pathways and their emission dynamics into the ocean using Lagrangian methods have been published [ 18 – 21 ]. However, few studies have focused on the spatiotemporal dynamics that drive debris, retention, and remobilization within riverine-estuarine environments. In particular, the role of geomorphological features (e.g. meanders and bank types) and transitional patterns between domains (e.g. tributary – main river – estuarine lagoon) are poorly understood. We address these gaps by estimating seasonal transport rates, identifying the main retention zones and their characteristics, assessing the role of an estuarine lagoon and the extent of debris export into the sea. In this paper, we present a Lagrangian approach by tracking floating drifters in the Umgeni River to represent the transport and fate of floating macroplastics. Our approach allowed for observations of short- and long-term transport patterns, which are challenging to capture in situ. The drifters were deployed between 2022–2023 across three different hydrological phases to capture the response of debris transport to dry and wet periods. The study analyses drifter movements in the Umgeni River to answer the following questions (1) what is the nature of transport rate (km day − 1 ), retention zones and in-river fate of floating macroplastics?; (2) what are the characteristics of the main retention zones in the catchment and how do they impact river-to-ocean emissions?; (3) what is the role of the estuarine lagoon on macroplastic transport; and (4) what is the coastal route of macroplastics exported to the ocean? We explore the influence of rainfall and river level on drifter transfer and retention to gain insights on the response of floating macroplastics transport to hydrometeorological drivers. The inclusion of fine-scale drifter dynamics and seasonality in understanding river plastic transport is essential in guiding river-to-ocean emission modelling and enhancing effective management strategies. 2.0 Results 2.1 Higher drifter transport in the rainy seasons We investigated the seasonal transfer of floating debris along the Umgeni River by exploring daily net and total drifter transport (km day − 1 ) across three hydrological phases. We observed an increasing trend in the median and interquartile range of net transport from phase A (dry period) towards phase B and C (peak and recessive wet period, respectively, Fig. 1 A). Due to high variance and many outliers, the means do not reflect this same trend through all three phases. Interestingly, the maximum drifter net transport event was observed during the dry period (Phase A) at 17.21 km day − 1 , this indicates high variability in drifter transport independent of the season. Peak drifter transport extremes were clustered, with eight of the ten highest transport events occurring only on two days, each in phases A and C. Both days coincided with river levels at 0.48 and 0.50 metres above the annual average of 2.63 metres, respectively. Kruskal–Wallis and pair–wise comparisons using the Mann–Whitney U–test both showed significant differences (Phase A: p = 6.61 x 10 − 6 ; Phase B: p = 8.81 x 10 − 11 and Phase C: p = 9.25 x 10 − 6 ) in seasonal median drifter net transport between the hydrological phases (Supplementary Information – Table S1 ). The drifters had high onset net transport in the beginning of each deployment month during all phases ranging 0.32–1.76 km day − 1 . This decreased during the subsequent months to a range of 0–0.36 km day − 1 , with intermittent peaks of high transport reaching up to 9 km day − 1 . The net-to-total drifter transport ratio ranged between 0.24 and 0.44 throughout the three phases. The small percentage of drifters that transitioned into the ocean exhibited high coastal transport rates (mean = 47 km day − 1 ), far exceeding the mean transport recorded within the river system (Supplementary Information – Figure S1 ). The four pilot devices we deployed prior to the full study captured the 1:50–100 year return flood on the Umgeni River of April 2022. One of the pilot devices was flushed 3.5 km downstream into the Indian Ocean within six hours (Figure S2). 2.2 Drifter Retention Episodes and Locations On average, drifters experienced one retention event every day across all hydrological phases, with maximum retention counts per day of 5, 4 and 5 times per day during phases A, B and C, respectively (Fig. 1 C). The daily retention counts exhibited no significant differences across all phases and between pairs of phases (Supplementary Information – Table S2). Mean retention durations were longer in the estuarine zone, ranging between 102 and 654 hours compared to the upstream river segment in range 47 and 509 hrs. We observe a notable decrease in the duration of retention episodes as we move from the dry period (phase A) towards the wet season (Phases B and C; Fig. 1 D). Drifters were retained for an average range of 40–508 hours per episode across all phases, with a notable reduction during phase C by a factor of 17 and 11 in comparison to phases A and B, respectively. Phase A exhibited the longest mean retention episode duration (mean = 508 hrs; median = 16 hrs), while Phase C recorded the lowest mean (mean = 40 hrs; median = 5hrs). The variability of retention durations was highest during phase A (STD = 1125 hrs), compared to phase B (STD = 748 hrs) and phase C (STD = 71 hrs). Significant differences in retention durations were only observed between hydrological phases A v. C and B v. C (Supplementary Information – Table S3). We characterised physical attributes that influence debris stranding by classfying retention habitats into three geomorphic zones: river meanders, straight channel sections and the estuary-lagoon For each retention zone, we catergorised the surface contact into rocky, vegetative, sandy/muddy, and infrastructure bank types (Fig. 2 ). Across all phases combined, river meanders dominated debris trapping, hosting 48% of retention events escpecially during the peak wet phase (B). Straight river channels hosted a combined 33% of retention episodes across all phases, with a peak also in the wettest phase (B). For riverbank types, vegetated river banks recorded the highest stranding episodes, accounting for 55% of all retention events. The majority of these episodes were observed during the peak wet phase (B). Sandy/muddy riverbanks exhibited 21% of all retention episodes and reached their high in the wettest phase. Overall, rocky banks and infrasructure had relatively low retention proportions. The meander zone vegetated bank type combination consistently contributed the highest number of drifter retention episodes during the entire study period. The second most frequent retention compartment was straight reaches with sandy/muddy banks, though these produced substantially fewer episodes than meander-vegetated banks. 2.3 Drifter Transition Probabilities and Fates We explored drifter fates (accumulation zones) and the likelihood of drifter transfer between domains across all hydrological phases. Overall, most drifters ended up within the riverine system at the end of the study period (45% – tributaries and main rivers) and a smaller percentage reached the estuary (23%) and Indian ocean (21%) (Fig. 3 A). Drifter fates showed seasonal variation, with a notable decrease in estuarian fates from phase A through C, whilst ocean bound drifters increased from phase A through C (Fig. 3 A). This suggests low drifter exports into the Indian ocean during low rainfall periods. In contrast, tributary and main river fates displayed an increasing trend from the dry phase A towards the wettest phase B, then finally dropped in the recessive phase C. Majority of drifters (85%) exported to the ocean beached within a 0.5 km radius north of the river mouth, while the other 15% (n = 2) drifted south for 6 km and 149 km. Interestingly, the peak wet phase – B had the most combined drifters that showed their last position within the riverine environment (tributaries and main rivers). The Umgeni River reflected low overall connectivity of drifter transfer between hydrological domains across all phases ranging 0–40% per domain pair (Fig. 3 B). For all phases, the combined transition of drifters from the main river to the estuary dominated with a probability of 80% followed by estuary to ocean transitions (40%) and lastly tributary to main river transfers were only 22%. Interestingly, drifter transitions from the main river to the estuary was high particularly during the dry phase A, while estuary to ocean transitions were highest during the recessive phase C. 2.4 Role of Rainfall and River Level on Drifter Transport and Retention The study year was particularly dry, with median daily rainfall ranging between 0–0.06 mm across all phases. Isolated heavy rainfall events dominated the total rainfall received during each hydrological phase (phase A = 86 mm; phase B = 500 mm and phase C = 329 mm). We applied regression and Pearson correlation to assess the influence of rainfall and river level on the spatiotemporal dynamics of drifter transport. These assessments included non–lagged to lagged (0–3 days) mean daily rainfall and river level observations. Linearity (coefficient of determination, R 2 = 0.32) and significant positive correlation (r 2 = 0.57; p-value = 6.61 x 10 − 4 ) was only observed between non-lagged rainfall and drifter transport during the moderate rainfall phase C (Fig. 4 ). Overall, discrete rainfall and river level observations, non–lagged and lagged (0–3 days) exhibited no linearity or significant correlation with drifter transport across all phases (Supplementary Information – Table S9). We further explored the effect of cumulative lagged rainfall and river level (1–3 days) on drifter transport, which also indicated no linearity or correlation across all phases. There was no statistically significant relationship in daily drifter retention counts with either daily rainfall or river level across all phases. Similarly, the duration of individual retention events indicated no statistically significant correlation with daily rainfall and river level changes. Similarly, the duration of retention episodes showed no correlation with rainfall and river level observations across all seasons. These results did not change significantly when applying 0–3 days lagged rainfall and river level data. The weak correlations suggest that retention of drifters within the Umgeni River might be influenced more by geomorphological features and in–channel infrastructure. 3.0 Discussion 3.1 Seasonal Variability in Drifter Transport Dynamics Besides a general trend of increased drifter transport during wet phases, we find that episodic hydrological events in any season play an important role in debris displacement. The peak daily transport event followed an elevated river level which was 18% higher than the annual mean, counterintuitively occurring during the dry rainfall phase (A). This reflects the role of high force hydrological conditions in driving macroplastic transport [ 16 , 23 , 24 ]. Such events have the capacity to flush debris stranded along banks and trapped at in–channel obstacles. Across all phases, mean daily transport remained low (0.31, 0.28 and 1.16 km day − 1 for phases A, B and C, respectively), indicating that debris moves in short displacement episodes rather than continuous transport. The observation that almost every drifter experienced a daily retention event (mean = 1 per day) is evidence of the predominant stranding of floating debris in the Umgeni River. This pattern aligns with observations from other rivers, where macroplastic transport was characterized by short distance stepwise displacements [ 18 , 19 , 25 ]. Similarly, studies have reported limited plastic transport under moderate flows with long distance trajectories occurring only during isolated heavy hydrological events [ 14 , 26 ]. The high onset drifter transport observed across all phases immediately after the devices were released, can be explained by the initial drifter deployment position in the middle of the river channel. This likely resulted in less impeding of drifters due to the absence of retention features (vegetation or rocky obstacles) and the momentum when releasing drifters could have slightly enhanced early transport velocity. However, this momentum drops as drifters encounter in–channel obstacles, hydraulic infrastructure and bank roughness down the river. The higher transport rates after drifters reach the ocean reflects the context of open ocean transport driven by wind gusts and surface currents in the absence of mechanical obstacles relative to channel constrained fluvial transport [ 27 ] and thus further provides a useful control for the in-river transport recording method. Our findings show that debris transport patterns in the Umgeni River are dominated by short-distance stepwise displacements and episodic large transport pulses driven by high river discharge. Even during the wet phases daily drifter transport remained low, indicating that debris transport is governed by multiple interactions between micro-scale hydrologic and geomorphologic factors [ 16 , 28 , 29 ]. Further studies should investigate integrating fine scale spatiotemporal observations of debris movement and hydrological models to better understand their interaction responses. This would help determine hydrometeorological thresholds that induce plastic movements. It is also worth noting that the study year’s relatively low seasonal rainfall might have further reduced transport contrasts induced by hydrologic forces between phases, muting potential relationships. 3.2 Seasonal Drifter Retention Dynamics and Accumulation Zones The absence of statistically significant differences in drifter retention counts between phases indicates that retention in the Umgeni River is likely stochastic with limited influence of seasonal hydrological changes. Such observations could suggest that local river micro–habitats (e.g. meanders, vegetative and rocky riverbanks) influence debris retentions more strongly than seasonal hydrologic variability [ 30 ]. The general reduction in drifter retention durations as we move from the dry season (phase A) towards the wet season (phases B and C) likely shows expected drifter mobilization and reduced residence time due to increased rainfall and river flows [ 23 , 29 ]. Interestingly, shorter retention durations during the recessive wet phase C compared to the peak wet phase B, could indicate that high water level during the peak wet phase might push debris to the upper banks which hold increased vegetative/rocky roughness and thus more likely lock debris compared to the lower and smoother bank portions. The domination of river meanders in retaining debris (48% of all retention episodes) shows how channel bends direct floating debris towards river bend banks. This increases the likelihood of debris contacting the bank, where enhanced friction due to bank roughness promote trapping. Meanders could, thus, be strategically selected for cleanup and interceptor barrier placement [ 31 ]. Increased retention events during the wet phase at meanders probably indicate that high discharges strengthen secondary flows and drive debris to the upper banks where it is locked in vegetation. This aligns with studies which report meanders as persistent debris accumulation zones due to the lateral shear [ 32 , 33 ]. While debris retentions at meanders and straight channels were high during the wettest phase (B), estuarine-lagoon retention episodes reached their high in the dry phase (A). This likely shows the effect of reduced river discharge during low flow conditions, which reduce debris flushing and enhance low energy lagoonal waters which promote debris trapping. During the dry phases, dampened flows allow for tidal oscillation at the lagoon where near-bank mangroves amplify debris stranding and thus prolong residence time. These observed patterns are consistent with the documented role of estuaries as debris sinks which transfer debris offshore only during high magnitude hydrological conditions [ 20 , 34 ]. Across all zones, vegetated banks were the leading retention surfaces, followed by sandy/muddy banks, both exhibiting peak retentions during the wettest phase B. As expected, vegetation increases surface roughness, reduces flow and physically impedes floating debris whilst sandy/muddy banks provide deposition zones [ 9 , 10 , 28 ]. Our results highlight that retention is the result of interactions between geomorphic setting, surface roughness and hydrologic forces. In the specific case of the Umgeni River, prolonged drifter stranding in the estuary is likely explained by the presence of a lagoon bordered with beach mangroves [ 35 ]. We did not consider wind and tidal effects which have been suggested to influence debris trapping too [ 18 , 29 ]. There is need to identify retention prone zones for targeted cleanups and monitoring. Drone based river flow visualizations could provide how the water interacts with vegetation, sand/muddy banks and infrastructure to influence debris retention at a finer spatial scale. 3.3 Drifter Pathways and Fates Most drifters remained in the riverine system upstream of the estuary-lagoon up until they ceased functioning. This reflects that a substantial proportion of debris is unlikely to reach the estuary in the first place, but whatever does enter the lagoon exhibits prolonged residence time there with limited ocean transfer potential. This supports the insight that macroplastics remain locked and lost in river systems for long periods [ 11 , 28 , 36 ]. The comparable seasonal changes in lagoon and ocean bound drifters, highlight that both riverine and estuarine environments provide a multi-stage retention system and indicate the potential flushing of debris from the estuary facilitated by increased rainfall and upstream river flows [ 19 , 37 ]. The increase in estuarine fates during the dry phase shows the enabling effect of reduced river flows in prolonging residence time and tidal trapping of debris in estuarine systems [ 21 , 35 , 38 ]. However, as river flows increase during the wet periods, dominant downstream river discharge enhance seaward debris transport [ 34 , 35 ]. The observed behaviour of ocean bound drifters to accumulate on nearshore banks reflect the role of ocean hydrodynamics that limit debris drifting further into the open sea [ 39 , 40 ]. The increase in tributary and main river drifter fates during the peak wet phase could support the hypothesis that high-flow conditions remobilize debris upstream but might not drive an increase in transfers to the estuary and ocean [ 9 , 10 ]. This results in debris being redistributed and re-trapped within the riverine ecosystem. The transition analysis reflects low interdomain connectivity of debris transport in the Umgeni catchment. Despite the high main river-to- estuary drifter transition rate (80%), the lower estuary-to-ocean transitions (40%) shows the strong attenuation within the estuarine zone. In addition to the role of stagnant lagoon waters, the attenuation is likely reinforced by tidal counter-currents and mangrove forests which border the lagoon banks. It is worth considering that the observed fates to some extent might also reflect battery depletion or loss of GPS signal connectivity. Some drifters may have continued downstream displacement despite battery depletion or signal loss. As a result, our findings present fates and transition rates between domains within our observation window. 3.4 Drifter Responses to Hydrologic Forcing In general, weak and non-significant correlations between rainfall or river level and drifter behaviour across phases indicate that simple linear relationships are inadequate to describe transport and retention dynamics. This finding aligns with reports of low correlations between rainfall or discharge with plastic transport in urban catchments [ 37 , 41 ]. Overall correlations between 0–3 days lagged rainfall or river level, and drifter transport were weak and statistically insignificant. This suggests that the relationship between rising river levels and drifter mobility is non-linear or threshold-governed. However, a strong and significant correlation between non–lagged rainfall and drifter transport was observed during the recessive wet phase C. It is worth noting that the study year was relatively dry compared to historical observations, with heavy erratic rainfall events contributing much to seasonal totals. In contrast to general seasonality, extreme rainfall events have been identified as strong drivers of macroplastic transport responsible for plastic flushing and long-distance movements [ 14 , 37 ]. Meanwhile, the lack of significant lagged rainfall and river level effects, could suggest buffering effects of river morphology – bank roughness leading to temporary storage and delayed release of debris [ 9 ]. This corresponds to similar observations in sediment transport studies, where sediment mobilization often lags behind rainfall events, reflecting delayed response to rainfall forcing [ 42 , 43 ]. Future research should answer if larger temporal scale or lags are required to represent the relationship between rainfall or river level and debris transport. It is to note that we had to interpolate eight months of river level data (see methods). This potentially reduced hydrologic accuracy. Additional drivers of riverine debris transport such as wind forcing, anthropogenic activities, controlled flood gate water release from the Inanda hydro dam in the Umgeni catchment and tidal forces may have influenced drifter movements aside of rainfall and river level. In the future, efforts should be directed towards integrating continuous fine scale debris monitoring and more hydrologic parameters at a comparable temporal scale to better understand these relationships. 3.5 Limitations of the Methodological Strategy While drifters provide valuable insights of debris transport behaviour, they represent only a portion of the actual macroplastics present in the river. Actual macroplastic debris have varying buoyancy, size, shape and material composition which influence floating stability and how the items interact with river flow and obstacles [ 44 , 45 ]. This is a limitation of the representativity of the devices towards the real world. The drifters provide uniform geometry and buoyancy which can result in smooth floating trajectories that underestimate retention and transport velocity of irregular plastic items [ 13 ]. Drifter effectiveness is high when the observed trajectories represent a range of buoyant plastics rather than full spectrum of river plastic debris. Our drifter deployment strategy and observation window might have missed longer–term storage and mobilization of debris. It is possible for debris to be buried in floodplains and trapped on banks for extended periods from days to years, leading to transport and retention dynamics beyond the timescale of the study [ 28 , 46 ]. It is important if future studies could deploy, multiple drifter design spanning a range of densities and shapes to improve representation of debris characteristics. 4.0 Methods 4.1 Study Area We conducted the study in the lower catchment of the South African Umgeni River. Specifically, on a 24 km stretch between the Inanda Dam and the north of Durban where the stream discharges into the Indian Ocean (Fig. 5 ). This segment of the river traverses through a densely urbanized region of Durban (population ~ 1.3 million), characterized by informal settlements, residential and industrial land use. The river network in this study includes four tributaries – the Umhlangane, Palmiet, Piesang and Aller Rivers, with the Umhlangane being the major tributary stretching 21 km through Durban’s largest high–density northern suburb of Kwamashu. The selected rivers have varying flow regimes with Strahler orders in the range of 7–9th. The estuary, which includes the Blue Lagoon, experiences semi–diurnal tidal cycles ranging from 0.2–2 meters. The annual average river flow rate for the Umgeni is 21 m 3 /s, which varies greatly and can reach extreme peaks of ~ 1300 m 3 /s for 100–year flood return periods [ 47 ]. Notably, the study period was preceded by floods which occurred in the catchment between 10th – 14th of April, with peak mean daily rainfall reaching 170.8 mm. The Umgeni catchment annual rainfall ranges from 410 to 1450 mm, with mean annual runoff in the range of 72–680 mm. Most rainfall occurs during the summer season (October–April) with monthly averages in the range 107–118 mm, influenced by the Indian Ocean Agulhas Current. Tropical cyclones induce floods which dominate the southern coastal region of South Africa. The abundant plastic pollution in this section of Umgeni River is driven by mismanaged waste from informal settlements and industrial areas mobilized and transported by dynamic river flow characterized by isolated floods [ 5 ]. 4.2 Drifter Assembly We employed floating GPS tracking devices approximating in size, shape and density floating macro–debris commonly found in the urbanized region of central Durban (e.g. bottles and food–packs) in alignment with [ 21 ]. The drifters operated using Global Navigation Satellite Systems (GNSS) for positioning and Global System for Mobile Communication (GSM) for data transmission. We programmed the devices to send to a server geo-location (coordinates) every five minutes when in motion and every two hours when stationary, to optimize both spatiotemporal resolution and battery life. Every drifter location update included the corresponding timestamp and speed of the drifter. Each drifter comprised a printed circuit board (PCB) with an antenna (LightBug Pro, Lightbug, Bristol, U.K), lithium battery (3.7 V; 20,000 mAh) vacuum sealed and placed in ~ 500 ml HDPE screw–top cylinders padded with Styrofoam. The container screw–tops were sealed off using silicon. We labeled the plastic containers with experiment details, contents, and contact information. On average, each drifter had a total density of 0.94 g/cm 3 (weight = ~ 472 g) with a height and diameter of ~ 8 cm and ~ 10 cm, respectively. 4.3 Drifter Spatial Deployment We deployed 66 drifters (+ 4 pilot devices preceding experiment) along the Umgeni River and five of its tributaries in three phases (22 per phase). Drifter deployment phases were defined based on rainfall received during the study period. The first deployment phase was carried out in June 2022 (phase A – dry phase), followed by the second in October 2022 (phase B – wet phase) and the last deployment in February 2023 (phase C – receding wet phase; Supplementary Information – Table S4). Prior to the full study we deployed four pilot drifters in the Umgeni River for testing, 18 km upstream of the river mouth. We distributed the drifters at equidistant river section lengths using the river mouth as the datum and for tributaries using the downstream confluence points as the datum. Tributaries with greater lengths and varying Strahler stream orders were prioritized for drifter deployments. This selection allowed clear observation of drifter movements over extended distances. The furthest inland deployment site was 27 km upstream of the river's mouth on the Piesang River. The estuary deployments – in the Blue Lagoon – were the closest to the ocean at 2 km. Estuarian sites were included to specifically investigate the transition potential from river to ocean influenced by the tides. Further considerations for the deployment locations were field team safety and site accessibility. Consequently, some deployment locations deviated from an arithmetically equidistant distribution along the river system. In total, we designated 13 drifter deployment locations across the Umgeni River study segment (Supplementary Information: Table S3). At each deployment location we released drifters in duplicates to enhance study robustness and account for the expected variability in river conditions. The drifters were preferentially released in the middle of the river cross section (width range: 11–125 meters) to avoid premature stranding on the riverbank (Supplementary Information: Table S5). 4.4 Processing of Drifter Spatiotemporal Data We processed drifter trajectories using QGIS and Geo Pandas Python Library. Only GNSS location records falling within the 100–yr floodplain zone was retained for the processing and analysis of drifter trajectories, while GSM and Wi–Fi triangulated positions were removed due to their spatial inaccuracy. The spatial constraint to the floodplain maximized the inclusion of only relevant data points, excluding e.g. artefactual data due to manipulation through human drifter displacement. We filtered the GNSS position updates to include only records with a reported accuracy range smaller than the distance moved between the current and previous drifter location. The average position accuracy was 8 meters. The deployment timestamp for each drifter was recorded on a field datasheet and assigned as the reference start time for its trajectory. This ensured consistent start points across deployment phases and eliminated the positional offsets introduced by GPS fixes at the start of each trajectory. 4.4.1 Calculating Drifter Transport We described drifter trajectories along the river system into two transport regimes, that is total and net drifter transport. Total drifter transport includes both longitudinal and lateral movements, while net transport describes only the longitudinal movement along the stream length. For net transport, drifter trajectory points were snapped to the river network centerline and segmented into daily trips of 24–hour intervals to standardize meaningful longitudinal transport units. The river centerline was represented by a continuous arc length parameterized curve \(C:\left[0,L\right]\to{\mathbb{R}}^{2}\) , where \(s\) denotes scalar distance(m) measured from an upstream origin along the river centerline and \(L\) is the total length of the centerline. The function \(C\left(s\right)=(x\left(s\right),y\left(s\right))\) returns the two-dimensional cartesian coordinates of the point located at distance \(s\) along the centerline. For each daily segment, the first \(\left({P}_{1}\right)\) and last \(\left({P}_{n}\right)\) recorded drifter positions were projected orthogonally onto the river centerline to clip their corresponding chainage coordinates: $${s}_{1}=arg{min}_{s\in\left[0,L\right]}\left|\right|{P}_{1}-C\left(s\right)\left|\right|,{s}_{n}=arg{min}_{s\in\left[0,L\right]}\left|\right|{P}_{n}-C\left(s\right)\left|\right|$$ 1 The daily net downstream transport (km day − 1 ) was then computed as: $${D}_{net,centerline}=\frac{\left|{s}_{n}-{s}_{1}\right|}{1000};\left({kmday}^{-1}\right)$$ 2 4.4.2 Calculating Drifter Retention and Fate A retention episode is defined as the time in which a drifter is stationary within the river system. We calculated drifter retention episodes based on the method by [ 21 ]. We smoothed the drifter velocity data using a 60–minute moving average and used a threshold of < 0.1m/s to define a retention episode. Retention frequency and duration were grouped based on the hydrological phases (A–C) to assess seasonal variability. We assessed the role of river geomorphology on the spatial distribution of retention episodes by identifying common retention zones along the river system (meander, straight stretches and estuary) using Esri satellite maps (2021–2022). Each retention zone was overlaid with four riverbank categories (i) rocky, (ii) vegetated, (iii) sandy/muddy, and (iv) in-channel infrastructure. In-channel infrastructure was included as a retaining type due to it's documented potential to retain debris [ 48 , 49 ]. The fate domain of each drifter was considered as the geographical compartment into which the last recorded drifter location fell (e.g. river, estuary, or ocean). 4.4.3 Statistical Analysis To examine the floating drifter's behavior along the Umgeni River system across hydrological seasons (Phases A, B and C), we used descriptive statistics measures (mean, median and standard deviation) to provide an overview of drifter transport and retention episodes under varying hydrological phases. Because the data did not meet normality assumptions, we used non-parametric Kruskal-Wallis tests to evaluate overall statistical differences in daily drifter net transport, frequency of retention episodes, and their duration across the three independent phases. In cases where significant differences existed (p < 0.05), we conducted pair-wise comparisons using the Mann-Whitney U tests to identify specific phases which differed from each other. The Mann-Whitney U tests were used subsequent as a post-hoc pairwise comparison since the Kruskal-Wallis tests assess group differences but not pairwise contrasts. The influence of rainfall and river level on macro–debris transfer dynamics was assessed using linear regression and Pearson correlation coefficients. To capture delayed hydrological responses, we evaluated lagged rainfall and river level effects at 0–3-day intervals. River level data had missing records during the period 2022-12-15 to 2023-08-15. These gaps were infilled using random forest regression, trained on periods with concurrent observations of daily river level and rainfall between 2021–2023. The model used nonlinear relationships between rainfall and river level to predict the missing river level measurements. All statistical analyses were conducted in Python (version 3.9), using SciPy, stats models, and scikit–posthocs. Declarations 7.0 Funding Statement This study was funded by the donors of The Ocean Cleanup and by Tito’s Handmade Vodka under the 3 Rivers 3 Years Research Program. Author Contribution TM and CT: Conceptualization, Methodology. TG: Data curation, Formal analysis, Writing - Original draft preparation. TG: Visualization, Investigation. TM and CT: Supervision. TG, CT and TM: Writing- Reviewing and Editing. Acknowledgement We thank Laurent Lebreton, The Ocean Cleanup, for guidance on the study; and Stijn Pinson, The Ocean Cleanup, for support with fieldwork during the pilot phase. This study was funded by the donors of The Ocean Cleanup and by Tito’s Handmade Vodka under the 3 Rivers 3 Years Research Program. Data Availability The raw GPS tracker datasets and processed data used in this study are available in [doi.org/10.6084/m9.figshare.31558594](https:/tocu.sharepoint.com/sites/research/Gedeelde%20documenten/02%20Global%20River%20Sources/R201%20PRJ48%203R3Y%20River%201%20Umgeni%20SA/05%20Reporting/Manuscripts/2026_GPS_SciRep/doi.org/10.6084/m9.figshare.31558594) . Rainfall and river level data were obtained from eThekwini Municipality and are available on [eThekwini Datafeeds](https:/www.bing.com/ck/a?!&&p=75c9b5ecd9d6a0e825daf8f8761f8c3397dcaf0e83bdeed35f2d098ff7cc099eJmltdHM9MTc3MjE1MDQwMA&ptn=3&ver=2&hsh=4&fclid=2bb31aa9-b5e3-63aa-123f-08eeb49a6275&psq=ethekwini+datafeed&u=a1aHR0cHM6Ly9kYXRhLmV0aGVrd2luaWZld3MuZHVyYmFuLw) . References Al-Zawaidah, H., Ravazzolo, D. & Friedrich, H. Macroplastics in rivers: Present knowledge, issues and challenges. Environmental Science: Processes and Impacts vol. 23 535–552 Preprint at (2021). https://doi.org/10.1039/d0em00517g Sonke, J. E., Koenig, A., Segur, T. & Yakovenko, N. Global Environmental Plastic Dispersal under OECD Policy Scenarios toward 2060 . Sci Adv 11 (2025). https://www.science.org Nyberg, B., Harris, P. T., Kane, I. & Maes, T. Leaving a plastic legacy: Current and future scenarios for mismanaged plastic waste in rivers. Science Total Environment 869 , (2023). Pinto, R. B. et al. Catchment scale assessment of macroplastic pollution in the Odaw river, Ghana. Mar Pollut Bull 198 , (2024). Gutsa, T., Trois, C., de Vries, R. & Mani, T. Wasted shores: Using drones to monitor the spatio-temporal evolution of debris accumulation hotspots on South Africa’s Umgeni River. Science Total Environment 955 , (2024). Mikusheva, M. et al. Floating Riverine Litter Flux to the White Sea: Seasonal Changes in Abundance and Composition. J Mar. Sci. Eng 11 , (2023). Schreyers, L. J. et al. River plastic transport affected by tidal dynamics. Hydrol. Earth Syst. Sci. 28 , 589–610 (2024). Meijer, L. J. J., Van Emmerik, T., Van Der Ent, R., Schmidt, C. & Lebreton, L. More than 1000 Rivers Account for 80% of Global Riverine Plastic Emissions into the Ocean . Sci Adv vol. 7 (2021). https://www.science.org Liro, M., van Emmerik, T., Wyzga, B., Liro, J. & Mikuś, P. Macroplastic storage and remobilization in rivers. Water (Switzerland) 12 , (2020). Schreyers, L. J. et al. River plastic transport and storage budget. Preprint at (2023). https://doi.org/10.21203/rs.3.rs-3743805/v1 van Emmerik, T., Mellink, Y., Hauk, R., Waldschläger, K. & Schreyers, L. Rivers as Plastic Reservoirs. Frontiers Water 3 , (2022). Mellink, Y., van Emmerik, T., Kooi, M., Laufkötter, C. & Niemann, H. The Plastic Pathfinder: A macroplastic transport and fate model for terrestrial environments. Front Environ. Sci 10 , (2022). Kuizenga, B., Tasseron, P. F., Wendt-Potthoff, K. & van Emmerik, T. H. M. From source to sea: Floating macroplastic transport along the Rhine river. Front Environ. Sci 11 , (2023). van Emmerik, T. et al. River plastic transport and deposition amplified by extreme flood. Nat. Water . 1 , 514–522 (2023). Pessenlehner, S., Gmeiner, P., Habersack, H. & Liedermann, M. Understanding the spatio-temporal behaviour of riverine plastic transport and its significance for flux determination: insights from direct measurements in the Austrian Danube River. Front Earth Sci. (Lausanne) 12 , (2024). Tobias, C. & Roebroek, J. The role of hydrometeorology in river plastic pollution. (2024). Wickramarachchi, C., Niven, R. K. & Kramer, M. Numerical plastic transport modelling in fluvial systems: Review and formulation of boundary conditions. Water Research vol. 273 Preprint at (2025). https://doi.org/10.1016/j.watres.2024.122947 Mani, T. et al. The tidal trap — Seasonal transport of floating macrodebris in the bi-directional Chao Phraya River network on the Gulf of Thailand. Mar Pollut Bull 212 , (2025). Tramoy, R. et al. Transfer dynamics of macroplastics in estuaries – New insights from the Seine estuary: Part 2. Short-term dynamics based on GPS-trackers. Mar. Pollut Bull. 160 , 111566 (2020). Ledieu, L. et al. Macroplastic transfer dynamics in the Loire estuary: Similarities and specificities with macrotidal estuaries. Mar. Pollut Bull. 182 , 114019 (2022). Mani, T. et al. Gaining new insights into macroplastic transport ‘hotlines’ and fine-scale retention-remobilisation using small floating high-resolution satellite drifters in the Chao Phraya River estuary of Bangkok. Environmental Pollution 320 , (2023). Tosi Robinson, D. et al. Land-based plastic leakage into the aquatic environment from municipal solid waste – Waste flow diagram applied to Tuy Hoa City, Phu Yen, Vietnam. Waste Manage. 186 , 226–235 (2024). van Emmerik, T. et al. Seine Plastic Debris Transport Tenfolded During Increased River Discharge. Front Mar. Sci 6 , (2019). Hauk, R. et al. Macroplastic deposition and flushing in the Meuse river following the July 2021 European floods. Environ. Res. Lett. 18 , 124025 (2023). Gu, J., Zhang, Y., Sui, Y. & Chen, S. Lagrangian study of floating debris transport around the Pearl River Estuary in summer. Mar Pollut Bull 211 , (2025). van Emmerik, T. et al. Hydrology as a Driver of Floating River Plastic Transport. Earths Future 10 , (2022). Lebreton, L., Egger, M. & Slat, B. A global mass budget for positively buoyant macroplastic debris in the ocean. Sci Rep 9 , (2019). Liro, M., Mikuś, P. & Wyżga, B. First insight into the macroplastic storage in a mountain river: The role of in-river vegetation cover, wood jams and channel morphology. Science Total Environment 838 , (2022). Mellink, Y. A. M., van Emmerik, T. H. M. & Mani, T. Wind- and rain-driven macroplastic mobilization and transport on land. Sci Rep 14 , (2024). González-Fernández, D. et al. Floating macrolitter leaked from Europe into the ocean. Nat. Sustain. 4 , 474–483 (2021). Leone, G. et al. Integrating Bayesian Belief Networks in a toolbox for decision support on plastic clean-up technologies in rivers and estuaries. Environ. Pollut. 296 , 118721 (2022). Vriend, P. et al. Rapid Assessment of Floating Macroplastic Transport in the Rhine. Front Mar. Sci 7 , (2020). Gurnell, A. M., Bertoldi, W. & Corenblit, D. Changing river channels: The roles of hydrological processes, plants and pioneer fluvial landforms in humid temperate, mixed load, gravel bed rivers. Earth Sci. Rev. 111 , 129–141 (2012). Cheung, P. K., Cheung, L. T. O. & Fok, L. Seasonal variation in the abundance of marine plastic debris in the estuary of a subtropical macro-scale drainage basin in South China. Sci. Total Environ. 562 , 658–665 (2016). Ivar do Sul, J. A., Costa, M. F., Silva-Cavalcanti, J. S. & Araújo, M. C. B. Plastic debris retention and exportation by a mangrove forest patch. Mar. Pollut Bull. 78 , 252 (2014). Weiss, L. et al. The Missing Ocean Plastic Sink: Gone with the Rivers . (2023). https://www.science.org Van Emmerik, T. H. M. The impact of floods on plastic pollution. Global Sustainability 7 , (2024). Tramoy, R., Gasperi, J., Colasse, L. & Tassin, B. Transfer dynamic of macroplastics in estuaries — New insights from the Seine estuary: Part 1. Long term dynamic based on date-prints on stranded debris. Mar Pollut Bull 152 , (2020). Van Sebille, E. et al. The physical oceanography of the transport of floating marine debris. Environmental Research Letters vol. 15 Preprint at (2020). https://doi.org/10.1088/1748-9326/ab6d7d Ryan, P. G. & Perold, V. Limited dispersal of riverine litter onto nearby beaches during rainfall events. Estuar Coast Shelf Sci 251 , (2021). Pinto, R. B. et al. Exploring plastic transport dynamics in the Odaw river, Ghana. Front Environ. Sci 11 , (2023). Baniya, B. et al. Rainfall erosivity and sediment dynamics in the Himalaya catchment during the Melamchi flood in Nepal. J. Mt. Sci. 20 , 2993–3009 (2023). Valyrakis, M., Gilja, G., Liu, D. & Latessa, G. Transport of Floating Plastics through the Fluvial Vector: The Impact of Riparian Zones. Water (Switzerland) 16 , (2024). Born, M. P., Junge, L. V., Brüll, C., Waldschläger, K. & Schüttrumpf, H. Terminal settling and rising velocity prediction of macroplastics: Medical face masks as newly emerged objects of concern. Science Total Environment 908 , (2024). Lebreton, L. et al. Plastics in Freshwater Bodies. Plast. Ocean. 199–225. https://doi.org/10.1002/9781119768432.ch7 (2022). Tramoy, R., Gasperi, J., Colasse, L. & Tassin, B. Transfer dynamic of macroplastics in estuaries New insights from the Seine estuary: Part 1. Long term dynamic based on date-prints on stranded debris. Mar. Pollut Bull. 152 , 110894 (2020). Mphefu, T., Nkhonjera, G. & Alowo, R. Evaluation of the probability density function of unpredictable flood events in the Mngeni River basin, South Africa. J. Water Clim. Change . 16 , 1695–1714 (2025). Lotcheris, R. A. et al. Plastic Does Not Simply Flow into the Sea: River Transport Dynamics Affected by Tides and Floating Plants . (2023). https://ssrn.com/abstract=4449742 Honingh, D. et al. Urban river water level increase through plastic waste accumulation at a rack structure. Front. Earth Sci. (Lausanne) . 8 , 28 (2020). Additional Declarations No competing interests reported. Supplementary Files 2026GutsaetalLagoonsasGatekeepersUmgeniSMv.1.1.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 31 Mar, 2026 Reviewers agreed at journal 16 Mar, 2026 Reviewers invited by journal 13 Mar, 2026 Editor invited by journal 12 Mar, 2026 Editor assigned by journal 09 Mar, 2026 Submission checks completed at journal 09 Mar, 2026 First submitted to journal 06 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9052696","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":612435133,"identity":"aadf9455-6f38-4e52-90a5-309cd01e5578","order_by":0,"name":"Tadiwanashe Gutsa","email":"","orcid":"","institution":"University of KwaZulu-Natal","correspondingAuthor":false,"prefix":"","firstName":"Tadiwanashe","middleName":"","lastName":"Gutsa","suffix":""},{"id":612435135,"identity":"e7e6dc26-b01d-49c6-bdb7-a9f1a0533da8","order_by":1,"name":"Cristina Trois","email":"","orcid":"","institution":"University of KwaZulu-Natal","correspondingAuthor":false,"prefix":"","firstName":"Cristina","middleName":"","lastName":"Trois","suffix":""},{"id":612435136,"identity":"11534d6a-4bb0-44e2-8025-f965c46d9e51","order_by":2,"name":"Thomas Mani","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/0lEQVRIiWNgGAWjYNACAxsGAxibj5BiHoiWNAm4FjbitDAcJkGLPQN34ueCgvN15uzNBz98qLkjz8be/Owx7w4GeX6xAzhs4d0sPcPgtoRlz7FkyRnHnhm28RwzN+Y9w2A4c3YCLi0bpHmAWgxu5Jgx87AdTmCTSDCT5m1jSDC4jVPL5t88BuckDO6/MWP+8w+oRf75N0JatgFtOQC0hceMmbENZAsPAVsO826z5jFIltxwJi1ZsrfvMNAvOWWSc89I4PQLe3vv5ts8f+z4DY4fPvjhx7fD8vzsx7dJvN1hI88vjV0LAzNWUcYGCezKcQPGBlJ1jIJRMApGwTAGAIkkUvEJ1SbsAAAAAElFTkSuQmCC","orcid":"","institution":"The Ocean Cleanup","correspondingAuthor":true,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Mani","suffix":""}],"badges":[],"createdAt":"2026-03-06 16:53:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9052696/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9052696/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105825672,"identity":"8019c362-7a57-4c24-bbaa-015feb6aaa4b","added_by":"auto","created_at":"2026-03-31 13:56:39","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":117058,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Drifters daily net transport (km day\u003csup\u003e-1\u003c/sup\u003e), (B) daily rainfall (mm) distribution across the three hydrological phases, (C) Distribution of the number of drifter retention episodes and (D) duration of retention episodes across three hydrological phases. The outliers on the y–axis were cut off at 10 km day\u003csup\u003e-1 \u003c/sup\u003e(A) and 10 mm (B) for better visual comparability of the boxplots. P–values between hydrological phases are shown with asterisks above the solid black lines connecting the pairs (** \u0026lt;0.01; * \u0026lt;0.05) and the means are marked with a “+”.\u003c/p\u003e","description":"","filename":"Picture1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/288122f1e4f8c488f18150e7.jpg"},{"id":105825512,"identity":"f3431037-af6a-44c5-a501-54bc14bf0587","added_by":"auto","created_at":"2026-03-31 13:56:24","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":169940,"visible":true,"origin":"","legend":"\u003cp\u003eSeasonal heatmaps for the number of drifter retention episodes across geomorphological compartments and riverbank types.\u003c/p\u003e","description":"","filename":"Picture2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/baa10de6af327db04805fd41.jpg"},{"id":105825685,"identity":"455338b9-67b6-4b4f-8ab7-f102acced6e5","added_by":"auto","created_at":"2026-03-31 13:56:46","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":100009,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Changes in drifter river domain fates and (B) transitions between river domains.\u003c/p\u003e","description":"","filename":"Picture3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/a8123d96517deafd2c60b297.jpg"},{"id":105825510,"identity":"333428d0-cd79-4f25-94f0-8d33ba0c57da","added_by":"auto","created_at":"2026-03-31 13:56:24","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":182256,"visible":true,"origin":"","legend":"\u003cp\u003eRegression analysis between non lagged mean daily rainfall and daily drifter transport during (A) all phases, (B) phase A, (C) phase B and (D) phase C. The shaded area represents the 95% confidence intervals.\u003c/p\u003e","description":"","filename":"Picture4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/06cb71077ca6042c13c57463.jpg"},{"id":105825509,"identity":"13f996c1-b2f6-4a21-b97b-015f3a5b41dc","added_by":"auto","created_at":"2026-03-31 13:56:23","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":215765,"visible":true,"origin":"","legend":"\u003cp\u003eUmgeni River water network with an overview of GPS drifter release sites as well as rain gauge and river level gauge stations.\u003c/p\u003e","description":"","filename":"Picture5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/25f248cc0a97914296500ef7.jpg"},{"id":105825840,"identity":"10be7aed-460e-4b35-b8be-35143e8b4239","added_by":"auto","created_at":"2026-03-31 13:57:33","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1605937,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/9cd6d2f8-7e10-4acc-bcc8-03a51a5a452e.pdf"},{"id":105825693,"identity":"b3178a84-918f-4f0f-9add-07a8dfddd9f2","added_by":"auto","created_at":"2026-03-31 13:56:48","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":936325,"visible":true,"origin":"","legend":"","description":"","filename":"2026GutsaetalLagoonsasGatekeepersUmgeniSMv.1.1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9052696/v1/b2353bf7a0440b4c585d70d8.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Lagoons as Ocean Gatekeepers? Seasonal Transport and Retention Dynamics of Floating Macro-debris in the Umgeni River, South Africa","fulltext":[{"header":"1.0 Introduction","content":"\u003cp\u003ePlastic pollution in waterbodies poses threats to human livelihoods, local economies as well as marine and freshwater ecosystems [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. An expected tripling of plastic use and waste by 2060 will likely exacerbate the leakage of mismanaged plastic into freshwater systems and oceans [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Urban River catchments in regions characterized by dense population and inadequate waste management infrastructure are especially prone to plastic pollution [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. These contexts coupled with seasonal weather and river hydrodynamics introduce complexity in understanding the transfer of waterborne macroplastic from rivers to the ocean [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Rivers play a dual role: on the one hand they are primary transport agents of litter to the sea (0.8\u0026ndash;2.7\u0026nbsp;million metric tonnes/year) [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] and on the other hand they act as temporary storage reservoirs [\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Recent papers investigate the transport of floating plastic debris through river networks with the interplay of river hydrodynamics [\u003cspan additionalcitationids=\"CR13\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. But the varying intricate river catchment characteristics and uncertain longitudinal displacement probability across different landscapes limits our current comprehensive understanding of river plastic transport.\u003c/p\u003e \u003cp\u003eRecent studies employ Eulerian or Lagragian methods to investigate riverine macroplastic transport processes. Eulerian methods provide valuable insights into localized transport rates and the extent of floating litter in rivers but are limited in capturing broader spatiotemporal aspects including transfer probability and fate [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. These limitations are particularly pronounced when considering the effects of seasonal, hydrological, geomorphological and tidal forces [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Lagrangian approaches offer high resolution tracking of floating debris in rivers, capturing key processes such as temporary retention in floodplains or riverbanks, upstream-to-downstream debris movement and debris transfer dynamics between aquatic domains e.g. river mouth and sea. Recently some pivotal steps towards unravelling plastic pollution pathways and their emission dynamics into the ocean using Lagrangian methods have been published [\u003cspan additionalcitationids=\"CR19 CR20\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. However, few studies have focused on the spatiotemporal dynamics that drive debris, retention, and remobilization within riverine-estuarine environments. In particular, the role of geomorphological features (e.g. meanders and bank types) and transitional patterns between domains (e.g. tributary \u0026ndash; main river \u0026ndash; estuarine lagoon) are poorly understood. We address these gaps by estimating seasonal transport rates, identifying the main retention zones and their characteristics, assessing the role of an estuarine lagoon and the extent of debris export into the sea.\u003c/p\u003e \u003cp\u003eIn this paper, we present a Lagrangian approach by tracking floating drifters in the Umgeni River to represent the transport and fate of floating macroplastics. Our approach allowed for observations of short- and long-term transport patterns, which are challenging to capture in situ. The drifters were deployed between 2022\u0026ndash;2023 across three different hydrological phases to capture the response of debris transport to dry and wet periods. The study analyses drifter movements in the Umgeni River to answer the following questions (1) what is the nature of transport rate (km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), retention zones and in-river fate of floating macroplastics?; (2) what are the characteristics of the main retention zones in the catchment and how do they impact river-to-ocean emissions?; (3) what is the role of the estuarine lagoon on macroplastic transport; and (4) what is the coastal route of macroplastics exported to the ocean? We explore the influence of rainfall and river level on drifter transfer and retention to gain insights on the response of floating macroplastics transport to hydrometeorological drivers. The inclusion of fine-scale drifter dynamics and seasonality in understanding river plastic transport is essential in guiding river-to-ocean emission modelling and enhancing effective management strategies.\u003c/p\u003e"},{"header":"2.0 Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Higher drifter transport in the rainy seasons\u003c/h2\u003e \u003cp\u003eWe investigated the seasonal transfer of floating debris along the Umgeni River by exploring daily net and total drifter transport (km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) across three hydrological phases. We observed an increasing trend in the median and interquartile range of net transport from phase A (dry period) towards phase B and C (peak and recessive wet period, respectively, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Due to high variance and many outliers, the means do not reflect this same trend through all three phases. Interestingly, the maximum drifter net transport event was observed during the dry period (Phase A) at 17.21 km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, this indicates high variability in drifter transport independent of the season. Peak drifter transport extremes were clustered, with eight of the ten highest transport events occurring only on two days, each in phases A and C. Both days coincided with river levels at 0.48 and 0.50 metres above the annual average of 2.63 metres, respectively. Kruskal\u0026ndash;Wallis and pair\u0026ndash;wise comparisons using the Mann\u0026ndash;Whitney U\u0026ndash;test both showed significant differences (Phase A: p\u0026thinsp;=\u0026thinsp;6.61 x 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e; Phase B: p\u0026thinsp;=\u0026thinsp;8.81 x 10\u003csup\u003e\u0026minus;\u0026thinsp;11\u003c/sup\u003eand Phase C: p\u0026thinsp;=\u0026thinsp;9.25 x 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e) in seasonal median drifter net transport between the hydrological phases (Supplementary Information \u0026ndash; Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The drifters had high onset net transport in the beginning of each deployment month during all phases ranging 0.32\u0026ndash;1.76 km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. This decreased during the subsequent months to a range of 0\u0026ndash;0.36 km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, with intermittent peaks of high transport reaching up to 9 km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. The net-to-total drifter transport ratio ranged between 0.24 and 0.44 throughout the three phases. The small percentage of drifters that transitioned into the ocean exhibited high coastal transport rates (mean\u0026thinsp;=\u0026thinsp;47 km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), far exceeding the mean transport recorded within the river system (Supplementary Information \u0026ndash; Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The four pilot devices we deployed prior to the full study captured the 1:50\u0026ndash;100 year return flood on the Umgeni River of April 2022. One of the pilot devices was flushed 3.5 km downstream into the Indian Ocean within six hours (Figure S2).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Drifter Retention Episodes and Locations\u003c/h2\u003e \u003cp\u003eOn average, drifters experienced one retention event every day across all hydrological phases, with maximum retention counts per day of 5, 4 and 5 times per day during phases A, B and C, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). The daily retention counts exhibited no significant differences across all phases and between pairs of phases (Supplementary Information \u0026ndash; Table S2). Mean retention durations were longer in the estuarine zone, ranging between 102 and 654 hours compared to the upstream river segment in range 47 and 509 hrs. We observe a notable decrease in the duration of retention episodes as we move from the dry period (phase A) towards the wet season (Phases B and C; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). Drifters were retained for an average range of 40\u0026ndash;508 hours per episode across all phases, with a notable reduction during phase C by a factor of 17 and 11 in comparison to phases A and B, respectively. Phase A exhibited the longest mean retention episode duration (mean\u0026thinsp;=\u0026thinsp;508 hrs; median\u0026thinsp;=\u0026thinsp;16 hrs), while Phase C recorded the lowest mean (mean\u0026thinsp;=\u0026thinsp;40 hrs; median\u0026thinsp;=\u0026thinsp;5hrs). The variability of retention durations was highest during phase A (STD\u0026thinsp;=\u0026thinsp;1125 hrs), compared to phase B (STD\u0026thinsp;=\u0026thinsp;748 hrs) and phase C (STD\u0026thinsp;=\u0026thinsp;71 hrs). Significant differences in retention durations were only observed between hydrological phases A v. C and B v. C (Supplementary Information \u0026ndash; Table S3).\u003c/p\u003e \u003cp\u003eWe characterised physical attributes that influence debris stranding by classfying retention habitats into three geomorphic zones: river meanders, straight channel sections and the estuary-lagoon For each retention zone, we catergorised the surface contact into rocky, vegetative, sandy/muddy, and infrastructure bank types (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Across all phases combined, river meanders dominated debris trapping, hosting 48% of retention events escpecially during the peak wet phase (B). Straight river channels hosted a combined 33% of retention episodes across all phases, with a peak also in the wettest phase (B). For riverbank types, vegetated river banks recorded the highest stranding episodes, accounting for 55% of all retention events. The majority of these episodes were observed during the peak wet phase (B). Sandy/muddy riverbanks exhibited 21% of all retention episodes and reached their high in the wettest phase. Overall, rocky banks and infrasructure had relatively low retention proportions. The meander zone vegetated bank type combination consistently contributed the highest number of drifter retention episodes during the entire study period. The second most frequent retention compartment was straight reaches with sandy/muddy banks, though these produced substantially fewer episodes than meander-vegetated banks.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Drifter Transition Probabilities and Fates\u003c/h2\u003e \u003cp\u003eWe explored drifter fates (accumulation zones) and the likelihood of drifter transfer between domains across all hydrological phases. Overall, most drifters ended up within the riverine system at the end of the study period (45% \u0026ndash; tributaries and main rivers) and a smaller percentage reached the estuary (23%) and Indian ocean (21%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Drifter fates showed seasonal variation, with a notable decrease in estuarian fates from phase A through C, whilst ocean bound drifters increased from phase A through C (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). This suggests low drifter exports into the Indian ocean during low rainfall periods. In contrast, tributary and main river fates displayed an increasing trend from the dry phase A towards the wettest phase B, then finally dropped in the recessive phase C. Majority of drifters (85%) exported to the ocean beached within a 0.5 km radius north of the river mouth, while the other 15% (n\u0026thinsp;=\u0026thinsp;2) drifted south for 6 km and 149 km. Interestingly, the peak wet phase \u0026ndash; B had the most combined drifters that showed their last position within the riverine environment (tributaries and main rivers). The Umgeni River reflected low overall connectivity of drifter transfer between hydrological domains across all phases ranging 0\u0026ndash;40% per domain pair (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). For all phases, the combined transition of drifters from the main river to the estuary dominated with a probability of 80% followed by estuary to ocean transitions (40%) and lastly tributary to main river transfers were only 22%. Interestingly, drifter transitions from the main river to the estuary was high particularly during the dry phase A, while estuary to ocean transitions were highest during the recessive phase C.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Role of Rainfall and River Level on Drifter Transport and Retention\u003c/h2\u003e \u003cp\u003eThe study year was particularly dry, with median daily rainfall ranging between 0\u0026ndash;0.06 mm across all phases. Isolated heavy rainfall events dominated the total rainfall received during each hydrological phase (phase A\u0026thinsp;=\u0026thinsp;86 mm; phase B\u0026thinsp;=\u0026thinsp;500 mm and phase C\u0026thinsp;=\u0026thinsp;329 mm). We applied regression and Pearson correlation to assess the influence of rainfall and river level on the spatiotemporal dynamics of drifter transport. These assessments included non\u0026ndash;lagged to lagged (0\u0026ndash;3 days) mean daily rainfall and river level observations. Linearity (coefficient of determination, R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.32) and significant positive correlation (r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.57; p-value\u0026thinsp;=\u0026thinsp;6.61 x 10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e) was only observed between non-lagged rainfall and drifter transport during the moderate rainfall phase C (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Overall, discrete rainfall and river level observations, non\u0026ndash;lagged and lagged (0\u0026ndash;3 days) exhibited no linearity or significant correlation with drifter transport across all phases (Supplementary Information \u0026ndash; Table S9). We further explored the effect of cumulative lagged rainfall and river level (1\u0026ndash;3 days) on drifter transport, which also indicated no linearity or correlation across all phases.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThere was no statistically significant relationship in daily drifter retention counts with either daily rainfall or river level across all phases. Similarly, the duration of individual retention events indicated no statistically significant correlation with daily rainfall and river level changes. Similarly, the duration of retention episodes showed no correlation with rainfall and river level observations across all seasons. These results did not change significantly when applying 0\u0026ndash;3 days lagged rainfall and river level data. The weak correlations suggest that retention of drifters within the Umgeni River might be influenced more by geomorphological features and in\u0026ndash;channel infrastructure.\u003c/p\u003e \u003c/div\u003e"},{"header":"3.0 Discussion","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Seasonal Variability in Drifter Transport Dynamics\u003c/h2\u003e \u003cp\u003eBesides a general trend of increased drifter transport during wet phases, we find that episodic hydrological events in any season play an important role in debris displacement. The peak daily transport event followed an elevated river level which was 18% higher than the annual mean, counterintuitively occurring during the dry rainfall phase (A). This reflects the role of high force hydrological conditions in driving macroplastic transport [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Such events have the capacity to flush debris stranded along banks and trapped at in\u0026ndash;channel obstacles. Across all phases, mean daily transport remained low (0.31, 0.28 and 1.16 km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for phases A, B and C, respectively), indicating that debris moves in short displacement episodes rather than continuous transport. The observation that almost every drifter experienced a daily retention event (mean\u0026thinsp;=\u0026thinsp;1 per day) is evidence of the predominant stranding of floating debris in the Umgeni River. This pattern aligns with observations from other rivers, where macroplastic transport was characterized by short distance stepwise displacements [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Similarly, studies have reported limited plastic transport under moderate flows with long distance trajectories occurring only during isolated heavy hydrological events [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. The high onset drifter transport observed across all phases immediately after the devices were released, can be explained by the initial drifter deployment position in the middle of the river channel. This likely resulted in less impeding of drifters due to the absence of retention features (vegetation or rocky obstacles) and the momentum when releasing drifters could have slightly enhanced early transport velocity. However, this momentum drops as drifters encounter in\u0026ndash;channel obstacles, hydraulic infrastructure and bank roughness down the river. The higher transport rates after drifters reach the ocean reflects the context of open ocean transport driven by wind gusts and surface currents in the absence of mechanical obstacles relative to channel constrained fluvial transport [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] and thus further provides a useful control for the in-river transport recording method.\u003c/p\u003e \u003cp\u003eOur findings show that debris transport patterns in the Umgeni River are dominated by short-distance stepwise displacements and episodic large transport pulses driven by high river discharge. Even during the wet phases daily drifter transport remained low, indicating that debris transport is governed by multiple interactions between micro-scale hydrologic and geomorphologic factors [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Further studies should investigate integrating fine scale spatiotemporal observations of debris movement and hydrological models to better understand their interaction responses. This would help determine hydrometeorological thresholds that induce plastic movements. It is also worth noting that the study year\u0026rsquo;s relatively low seasonal rainfall might have further reduced transport contrasts induced by hydrologic forces between phases, muting potential relationships.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Seasonal Drifter Retention Dynamics and Accumulation Zones\u003c/h2\u003e \u003cp\u003eThe absence of statistically significant differences in drifter retention counts between phases indicates that retention in the Umgeni River is likely stochastic with limited influence of seasonal hydrological changes. Such observations could suggest that local river micro\u0026ndash;habitats (e.g. meanders, vegetative and rocky riverbanks) influence debris retentions more strongly than seasonal hydrologic variability [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The general reduction in drifter retention durations as we move from the dry season (phase A) towards the wet season (phases B and C) likely shows expected drifter mobilization and reduced residence time due to increased rainfall and river flows [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Interestingly, shorter retention durations during the recessive wet phase C compared to the peak wet phase B, could indicate that high water level during the peak wet phase might push debris to the upper banks which hold increased vegetative/rocky roughness and thus more likely lock debris compared to the lower and smoother bank portions.\u003c/p\u003e \u003cp\u003eThe domination of river meanders in retaining debris (48% of all retention episodes) shows how channel bends direct floating debris towards river bend banks. This increases the likelihood of debris contacting the bank, where enhanced friction due to bank roughness promote trapping. Meanders could, thus, be strategically selected for cleanup and interceptor barrier placement [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Increased retention events during the wet phase at meanders probably indicate that high discharges strengthen secondary flows and drive debris to the upper banks where it is locked in vegetation. This aligns with studies which report meanders as persistent debris accumulation zones due to the lateral shear [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. While debris retentions at meanders and straight channels were high during the wettest phase (B), estuarine-lagoon retention episodes reached their high in the dry phase (A). This likely shows the effect of reduced river discharge during low flow conditions, which reduce debris flushing and enhance low energy lagoonal waters which promote debris trapping. During the dry phases, dampened flows allow for tidal oscillation at the lagoon where near-bank mangroves amplify debris stranding and thus prolong residence time. These observed patterns are consistent with the documented role of estuaries as debris sinks which transfer debris offshore only during high magnitude hydrological conditions [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Across all zones, vegetated banks were the leading retention surfaces, followed by sandy/muddy banks, both exhibiting peak retentions during the wettest phase B. As expected, vegetation increases surface roughness, reduces flow and physically impedes floating debris whilst sandy/muddy banks provide deposition zones [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOur results highlight that retention is the result of interactions between geomorphic setting, surface roughness and hydrologic forces. In the specific case of the Umgeni River, prolonged drifter stranding in the estuary is likely explained by the presence of a lagoon bordered with beach mangroves [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. We did not consider wind and tidal effects which have been suggested to influence debris trapping too [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. There is need to identify retention prone zones for targeted cleanups and monitoring. Drone based river flow visualizations could provide how the water interacts with vegetation, sand/muddy banks and infrastructure to influence debris retention at a finer spatial scale.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Drifter Pathways and Fates\u003c/h2\u003e \u003cp\u003eMost drifters remained in the riverine system upstream of the estuary-lagoon up until they ceased functioning. This reflects that a substantial proportion of debris is unlikely to reach the estuary in the first place, but whatever does enter the lagoon exhibits prolonged residence time there with limited ocean transfer potential. This supports the insight that macroplastics remain locked and lost in river systems for long periods [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. The comparable seasonal changes in lagoon and ocean bound drifters, highlight that both riverine and estuarine environments provide a multi-stage retention system and indicate the potential flushing of debris from the estuary facilitated by increased rainfall and upstream river flows [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. The increase in estuarine fates during the dry phase shows the enabling effect of reduced river flows in prolonging residence time and tidal trapping of debris in estuarine systems [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. However, as river flows increase during the wet periods, dominant downstream river discharge enhance seaward debris transport [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. The observed behaviour of ocean bound drifters to accumulate on nearshore banks reflect the role of ocean hydrodynamics that limit debris drifting further into the open sea [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. The increase in tributary and main river drifter fates during the peak wet phase could support the hypothesis that high-flow conditions remobilize debris upstream but might not drive an increase in transfers to the estuary and ocean [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. This results in debris being redistributed and re-trapped within the riverine ecosystem. The transition analysis reflects low interdomain connectivity of debris transport in the Umgeni catchment. Despite the high main river-to- estuary drifter transition rate (80%), the lower estuary-to-ocean transitions (40%) shows the strong attenuation within the estuarine zone. In addition to the role of stagnant lagoon waters, the attenuation is likely reinforced by tidal counter-currents and mangrove forests which border the lagoon banks.\u003c/p\u003e \u003cp\u003eIt is worth considering that the observed fates to some extent might also reflect battery depletion or loss of GPS signal connectivity. Some drifters may have continued downstream displacement despite battery depletion or signal loss. As a result, our findings present fates and transition rates between domains within our observation window.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Drifter Responses to Hydrologic Forcing\u003c/h2\u003e \u003cp\u003eIn general, weak and non-significant correlations between rainfall or river level and drifter behaviour across phases indicate that simple linear relationships are inadequate to describe transport and retention dynamics. This finding aligns with reports of low correlations between rainfall or discharge with plastic transport in urban catchments [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Overall correlations between 0\u0026ndash;3 days lagged rainfall or river level, and drifter transport were weak and statistically insignificant. This suggests that the relationship between rising river levels and drifter mobility is non-linear or threshold-governed. However, a strong and significant correlation between non\u0026ndash;lagged rainfall and drifter transport was observed during the recessive wet phase C. It is worth noting that the study year was relatively dry compared to historical observations, with heavy erratic rainfall events contributing much to seasonal totals. In contrast to general seasonality, extreme rainfall events have been identified as strong drivers of macroplastic transport responsible for plastic flushing and long-distance movements [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Meanwhile, the lack of significant lagged rainfall and river level effects, could suggest buffering effects of river morphology \u0026ndash; bank roughness leading to temporary storage and delayed release of debris [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. This corresponds to similar observations in sediment transport studies, where sediment mobilization often lags behind rainfall events, reflecting delayed response to rainfall forcing [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFuture research should answer if larger temporal scale or lags are required to represent the relationship between rainfall or river level and debris transport. It is to note that we had to interpolate eight months of river level data (see methods). This potentially reduced hydrologic accuracy. Additional drivers of riverine debris transport such as wind forcing, anthropogenic activities, controlled flood gate water release from the Inanda hydro dam in the Umgeni catchment and tidal forces may have influenced drifter movements aside of rainfall and river level. In the future, efforts should be directed towards integrating continuous fine scale debris monitoring and more hydrologic parameters at a comparable temporal scale to better understand these relationships.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Limitations of the Methodological Strategy\u003c/h2\u003e \u003cp\u003eWhile drifters provide valuable insights of debris transport behaviour, they represent only a portion of the actual macroplastics present in the river. Actual macroplastic debris have varying buoyancy, size, shape and material composition which influence floating stability and how the items interact with river flow and obstacles [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. This is a limitation of the representativity of the devices towards the real world. The drifters provide uniform geometry and buoyancy which can result in smooth floating trajectories that underestimate retention and transport velocity of irregular plastic items [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDrifter effectiveness is high when the observed trajectories represent a range of buoyant plastics rather than full spectrum of river plastic debris. Our drifter deployment strategy and observation window might have missed longer\u0026ndash;term storage and mobilization of debris. It is possible for debris to be buried in floodplains and trapped on banks for extended periods from days to years, leading to transport and retention dynamics beyond the timescale of the study [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. It is important if future studies could deploy, multiple drifter design spanning a range of densities and shapes to improve representation of debris characteristics.\u003c/p\u003e \u003c/div\u003e"},{"header":"4.0 Methods","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Study Area\u003c/h2\u003e \u003cp\u003eWe conducted the study in the lower catchment of the South African Umgeni River. Specifically, on a 24 km stretch between the Inanda Dam and the north of Durban where the stream discharges into the Indian Ocean (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This segment of the river traverses through a densely urbanized region of Durban (population\u0026thinsp;~\u0026thinsp;1.3\u0026nbsp;million), characterized by informal settlements, residential and industrial land use. The river network in this study includes four tributaries \u0026ndash; the Umhlangane, Palmiet, Piesang and Aller Rivers, with the Umhlangane being the major tributary stretching 21 km through Durban\u0026rsquo;s largest high\u0026ndash;density northern suburb of Kwamashu. The selected rivers have varying flow regimes with Strahler orders in the range of 7\u0026ndash;9th. The estuary, which includes the Blue Lagoon, experiences semi\u0026ndash;diurnal tidal cycles ranging from 0.2\u0026ndash;2 meters. The annual average river flow rate for the Umgeni is 21 m\u003csup\u003e3\u003c/sup\u003e/s, which varies greatly and can reach extreme peaks of ~\u0026thinsp;1300 m\u003csup\u003e3\u003c/sup\u003e/s for 100\u0026ndash;year flood return periods [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Notably, the study period was preceded by floods which occurred in the catchment between 10th \u0026ndash; 14th of April, with peak mean daily rainfall reaching 170.8 mm. The Umgeni catchment annual rainfall ranges from 410 to 1450 mm, with mean annual runoff in the range of 72\u0026ndash;680 mm. Most rainfall occurs during the summer season (October\u0026ndash;April) with monthly averages in the range 107\u0026ndash;118 mm, influenced by the Indian Ocean Agulhas Current. Tropical cyclones induce floods which dominate the southern coastal region of South Africa. The abundant plastic pollution in this section of Umgeni River is driven by mismanaged waste from informal settlements and industrial areas mobilized and transported by dynamic river flow characterized by isolated floods [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Drifter Assembly\u003c/h2\u003e \u003cp\u003eWe employed floating GPS tracking devices approximating in size, shape and density floating macro\u0026ndash;debris commonly found in the urbanized region of central Durban (e.g. bottles and food\u0026ndash;packs) in alignment with [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. The drifters operated using Global Navigation Satellite Systems (GNSS) for positioning and Global System for Mobile Communication (GSM) for data transmission. We programmed the devices to send to a server geo-location (coordinates) every five minutes when in motion and every two hours when stationary, to optimize both spatiotemporal resolution and battery life. Every drifter location update included the corresponding timestamp and speed of the drifter. Each drifter comprised a printed circuit board (PCB) with an antenna (LightBug Pro, Lightbug, Bristol, U.K), lithium battery (3.7 V; 20,000 mAh) vacuum sealed and placed in ~\u0026thinsp;500 ml HDPE screw\u0026ndash;top cylinders padded with Styrofoam. The container screw\u0026ndash;tops were sealed off using silicon. We labeled the plastic containers with experiment details, contents, and contact information. On average, each drifter had a total density of 0.94 g/cm\u003csup\u003e3\u003c/sup\u003e (weight\u0026thinsp;=\u0026thinsp;~\u0026thinsp;472 g) with a height and diameter of ~\u0026thinsp;8 cm and ~\u0026thinsp;10 cm, respectively.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Drifter Spatial Deployment\u003c/h2\u003e \u003cp\u003eWe deployed 66 drifters (+\u0026thinsp;4 pilot devices preceding experiment) along the Umgeni River and five of its tributaries in three phases (22 per phase). Drifter deployment phases were defined based on rainfall received during the study period. The first deployment phase was carried out in June 2022 (phase A \u0026ndash; dry phase), followed by the second in October 2022 (phase B \u0026ndash; wet phase) and the last deployment in February 2023 (phase C \u0026ndash; receding wet phase; Supplementary Information \u0026ndash; Table S4). Prior to the full study we deployed four pilot drifters in the Umgeni River for testing, 18 km upstream of the river mouth. We distributed the drifters at equidistant river section lengths using the river mouth as the datum and for tributaries using the downstream confluence points as the datum. Tributaries with greater lengths and varying Strahler stream orders were prioritized for drifter deployments. This selection allowed clear observation of drifter movements over extended distances. The furthest inland deployment site was 27 km upstream of the river's mouth on the Piesang River. The estuary deployments \u0026ndash; in the Blue Lagoon \u0026ndash; were the closest to the ocean at 2 km. Estuarian sites were included to specifically investigate the transition potential from river to ocean influenced by the tides. Further considerations for the deployment locations were field team safety and site accessibility. Consequently, some deployment locations deviated from an arithmetically equidistant distribution along the river system. In total, we designated 13 drifter deployment locations across the Umgeni River study segment (Supplementary Information: Table S3). At each deployment location we released drifters in duplicates to enhance study robustness and account for the expected variability in river conditions. The drifters were preferentially released in the middle of the river cross section (width range: 11\u0026ndash;125 meters) to avoid premature stranding on the riverbank (Supplementary Information: Table S5).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Processing of Drifter Spatiotemporal Data\u003c/h2\u003e \u003cp\u003eWe processed drifter trajectories using QGIS and Geo Pandas Python Library. Only GNSS location records falling within the 100\u0026ndash;yr floodplain zone was retained for the processing and analysis of drifter trajectories, while GSM and Wi\u0026ndash;Fi triangulated positions were removed due to their spatial inaccuracy. The spatial constraint to the floodplain maximized the inclusion of only relevant data points, excluding e.g. artefactual data due to manipulation through human drifter displacement. We filtered the GNSS position updates to include only records with a reported accuracy range smaller than the distance moved between the current and previous drifter location. The average position accuracy was 8 meters. The deployment timestamp for each drifter was recorded on a field datasheet and assigned as the reference start time for its trajectory. This ensured consistent start points across deployment phases and eliminated the positional offsets introduced by GPS fixes at the start of each trajectory.\u003c/p\u003e \u003cdiv id=\"Sec18\" class=\"Section3\"\u003e \u003ch2\u003e4.4.1 Calculating Drifter Transport\u003c/h2\u003e \u003cp\u003eWe described drifter trajectories along the river system into two transport regimes, that is total and net drifter transport. Total drifter transport includes both longitudinal and lateral movements, while net transport describes only the longitudinal movement along the stream length. For net transport, drifter trajectory points were snapped to the river network centerline and segmented into daily trips of 24\u0026ndash;hour intervals to standardize meaningful longitudinal transport units. The river centerline was represented by a continuous arc length parameterized curve \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(C:\\left[0,L\\right]\\to{\\mathbb{R}}^{2}\\)\u003c/span\u003e\u003c/span\u003e, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(s\\)\u003c/span\u003e\u003c/span\u003e denotes scalar distance(m) measured from an upstream origin along the river centerline and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(L\\)\u003c/span\u003e\u003c/span\u003e is the total length of the centerline. The function \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(C\\left(s\\right)=(x\\left(s\\right),y\\left(s\\right))\\)\u003c/span\u003e\u003c/span\u003e returns the two-dimensional cartesian coordinates of the point located at distance \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(s\\)\u003c/span\u003e\u003c/span\u003e along the centerline. For each daily segment, the first \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\left({P}_{1}\\right)\\)\u003c/span\u003e\u003c/span\u003e and last \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\left({P}_{n}\\right)\\)\u003c/span\u003e\u003c/span\u003e recorded drifter positions were projected orthogonally onto the river centerline to clip their corresponding chainage coordinates:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$${s}_{1}=arg{min}_{s\\in\\left[0,L\\right]}\\left|\\right|{P}_{1}-C\\left(s\\right)\\left|\\right|,{s}_{n}=arg{min}_{s\\in\\left[0,L\\right]}\\left|\\right|{P}_{n}-C\\left(s\\right)\\left|\\right|$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe daily net downstream transport (km day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) was then computed as:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$${D}_{net,centerline}=\\frac{\\left|{s}_{n}-{s}_{1}\\right|}{1000};\\left({kmday}^{-1}\\right)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section3\"\u003e \u003ch2\u003e4.4.2 Calculating Drifter Retention and Fate\u003c/h2\u003e \u003cp\u003eA retention episode is defined as the time in which a drifter is stationary within the river system. We calculated drifter retention episodes based on the method by [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. We smoothed the drifter velocity data using a 60\u0026ndash;minute moving average and used a threshold of \u0026lt;\u0026thinsp;0.1m/s to define a retention episode. Retention frequency and duration were grouped based on the hydrological phases (A\u0026ndash;C) to assess seasonal variability. We assessed the role of river geomorphology on the spatial distribution of retention episodes by identifying common retention zones along the river system (meander, straight stretches and estuary) using Esri satellite maps (2021\u0026ndash;2022). Each retention zone was overlaid with four riverbank categories (i) rocky, (ii) vegetated, (iii) sandy/muddy, and (iv) in-channel infrastructure. In-channel infrastructure was included as a retaining type due to it's documented potential to retain debris [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. The fate domain of each drifter was considered as the geographical compartment into which the last recorded drifter location fell (e.g. river, estuary, or ocean).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section3\"\u003e \u003ch2\u003e4.4.3 Statistical Analysis\u003c/h2\u003e \u003cp\u003eTo examine the floating drifter's behavior along the Umgeni River system across hydrological seasons (Phases A, B and C), we used descriptive statistics measures (mean, median and standard deviation) to provide an overview of drifter transport and retention episodes under varying hydrological phases. Because the data did not meet normality assumptions, we used non-parametric Kruskal-Wallis tests to evaluate overall statistical differences in daily drifter net transport, frequency of retention episodes, and their duration across the three independent phases. In cases where significant differences existed (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), we conducted pair-wise comparisons using the Mann-Whitney U tests to identify specific phases which differed from each other. The Mann-Whitney U tests were used subsequent as a post-hoc pairwise comparison since the Kruskal-Wallis tests assess group differences but not pairwise contrasts. The influence of rainfall and river level on macro\u0026ndash;debris transfer dynamics was assessed using linear regression and Pearson correlation coefficients. To capture delayed hydrological responses, we evaluated lagged rainfall and river level effects at 0\u0026ndash;3-day intervals. River level data had missing records during the period 2022-12-15 to 2023-08-15. These gaps were infilled using random forest regression, trained on periods with concurrent observations of daily river level and rainfall between 2021\u0026ndash;2023. The model used nonlinear relationships between rainfall and river level to predict the missing river level measurements. All statistical analyses were conducted in Python (version 3.9), using SciPy, stats models, and scikit\u0026ndash;posthocs.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003e7.0 Funding Statement\u003c/h2\u003e \u003cp\u003eThis study was funded by the donors of The Ocean Cleanup and by Tito\u0026rsquo;s Handmade Vodka under the 3 Rivers 3 Years Research Program.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eTM and CT: Conceptualization, Methodology. TG: Data curation, Formal analysis, Writing - Original draft preparation. TG: Visualization, Investigation. TM and CT: Supervision. TG, CT and TM: Writing- Reviewing and Editing.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe thank Laurent Lebreton, The Ocean Cleanup, for guidance on the study; and Stijn Pinson, The Ocean Cleanup, for support with fieldwork during the pilot phase. This study was funded by the donors of The Ocean Cleanup and by Tito\u0026rsquo;s Handmade Vodka under the 3 Rivers 3 Years Research Program.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe raw GPS tracker datasets and processed data used in this study are available in [doi.org/10.6084/m9.figshare.31558594](https:/tocu.sharepoint.com/sites/research/Gedeelde%20documenten/02%20Global%20River%20Sources/R201%20PRJ48%203R3Y%20River%201%20Umgeni%20SA/05%20Reporting/Manuscripts/2026_GPS_SciRep/doi.org/10.6084/m9.figshare.31558594) . Rainfall and river level data were obtained from eThekwini Municipality and are available on [eThekwini Datafeeds](https:/www.bing.com/ck/a?!\u0026amp;\u0026amp;p=75c9b5ecd9d6a0e825daf8f8761f8c3397dcaf0e83bdeed35f2d098ff7cc099eJmltdHM9MTc3MjE1MDQwMA\u0026amp;ptn=3\u0026amp;ver=2\u0026amp;hsh=4\u0026amp;fclid=2bb31aa9-b5e3-63aa-123f-08eeb49a6275\u0026amp;psq=ethekwini+datafeed\u0026amp;u=a1aHR0cHM6Ly9kYXRhLmV0aGVrd2luaWZld3MuZHVyYmFuLw) .\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAl-Zawaidah, H., Ravazzolo, D. \u0026amp; Friedrich, H. Macroplastics in rivers: Present knowledge, issues and challenges. \u003cem\u003eEnvironmental Science: Processes and Impacts\u003c/em\u003e vol. 23 535\u0026ndash;552 Preprint at (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1039/d0em00517g\u003c/span\u003e\u003cspan address=\"10.1039/d0em00517g\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSonke, J. E., Koenig, A., Segur, T. \u0026amp; Yakovenko, N. \u003cem\u003eGlobal Environmental Plastic Dispersal under OECD Policy Scenarios toward 2060\u003c/em\u003e. \u003cem\u003eSci Adv\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.science.org\u003c/span\u003e\u003cspan address=\"https://www.science.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNyberg, B., Harris, P. T., Kane, I. \u0026amp; Maes, T. Leaving a plastic legacy: Current and future scenarios for mismanaged plastic waste in rivers. \u003cem\u003eScience Total Environment\u003c/em\u003e \u003cb\u003e869\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinto, R. B. et al. Catchment scale assessment of macroplastic pollution in the Odaw river, Ghana. \u003cem\u003eMar Pollut Bull\u003c/em\u003e \u003cb\u003e198\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGutsa, T., Trois, C., de Vries, R. \u0026amp; Mani, T. Wasted shores: Using drones to monitor the spatio-temporal evolution of debris accumulation hotspots on South Africa\u0026rsquo;s Umgeni River. \u003cem\u003eScience Total Environment\u003c/em\u003e \u003cb\u003e955\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMikusheva, M. et al. Floating Riverine Litter Flux to the White Sea: Seasonal Changes in Abundance and Composition. \u003cem\u003eJ Mar. Sci. Eng\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchreyers, L. J. et al. River plastic transport affected by tidal dynamics. \u003cem\u003eHydrol. Earth Syst. Sci.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 589\u0026ndash;610 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeijer, L. J. J., Van Emmerik, T., Van Der Ent, R., Schmidt, C. \u0026amp; Lebreton, L. \u003cem\u003eMore than 1000 Rivers Account for 80% of Global Riverine Plastic Emissions into the Ocean\u003c/em\u003e. \u003cem\u003eSci Adv\u003c/em\u003e vol. 7 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.science.org\u003c/span\u003e\u003cspan address=\"https://www.science.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiro, M., van Emmerik, T., Wyzga, B., Liro, J. \u0026amp; Mikuś, P. Macroplastic storage and remobilization in rivers. \u003cem\u003eWater (Switzerland)\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchreyers, L. J. et al. River plastic transport and storage budget. Preprint at (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21203/rs.3.rs-3743805/v1\u003c/span\u003e\u003cspan address=\"10.21203/rs.3.rs-3743805/v1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Emmerik, T., Mellink, Y., Hauk, R., Waldschl\u0026auml;ger, K. \u0026amp; Schreyers, L. Rivers as Plastic Reservoirs. \u003cem\u003eFrontiers Water\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMellink, Y., van Emmerik, T., Kooi, M., Laufk\u0026ouml;tter, C. \u0026amp; Niemann, H. The Plastic Pathfinder: A macroplastic transport and fate model for terrestrial environments. \u003cem\u003eFront Environ. Sci\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuizenga, B., Tasseron, P. F., Wendt-Potthoff, K. \u0026amp; van Emmerik, T. H. M. From source to sea: Floating macroplastic transport along the Rhine river. \u003cem\u003eFront Environ. Sci\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Emmerik, T. et al. River plastic transport and deposition amplified by extreme flood. \u003cem\u003eNat. Water\u003c/em\u003e. \u003cb\u003e1\u003c/b\u003e, 514\u0026ndash;522 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePessenlehner, S., Gmeiner, P., Habersack, H. \u0026amp; Liedermann, M. Understanding the spatio-temporal behaviour of riverine plastic transport and its significance for flux determination: insights from direct measurements in the Austrian Danube River. \u003cem\u003eFront Earth Sci. (Lausanne)\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTobias, C. \u0026amp; Roebroek, J. The role of hydrometeorology in river plastic pollution. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWickramarachchi, C., Niven, R. K. \u0026amp; Kramer, M. Numerical plastic transport modelling in fluvial systems: Review and formulation of boundary conditions. \u003cem\u003eWater Research\u003c/em\u003e vol. 273 Preprint at (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.watres.2024.122947\u003c/span\u003e\u003cspan address=\"10.1016/j.watres.2024.122947\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMani, T. et al. The tidal trap \u0026mdash; Seasonal transport of floating macrodebris in the bi-directional Chao Phraya River network on the Gulf of Thailand. \u003cem\u003eMar Pollut Bull\u003c/em\u003e \u003cb\u003e212\u003c/b\u003e, (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTramoy, R. et al. Transfer dynamics of macroplastics in estuaries \u0026ndash; New insights from the Seine estuary: Part 2. Short-term dynamics based on GPS-trackers. \u003cem\u003eMar. Pollut Bull.\u003c/em\u003e \u003cb\u003e160\u003c/b\u003e, 111566 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLedieu, L. et al. Macroplastic transfer dynamics in the Loire estuary: Similarities and specificities with macrotidal estuaries. \u003cem\u003eMar. Pollut Bull.\u003c/em\u003e \u003cb\u003e182\u003c/b\u003e, 114019 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMani, T. et al. Gaining new insights into macroplastic transport \u0026lsquo;hotlines\u0026rsquo; and fine-scale retention-remobilisation using small floating high-resolution satellite drifters in the Chao Phraya River estuary of Bangkok. \u003cem\u003eEnvironmental Pollution\u003c/em\u003e \u003cb\u003e320\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTosi Robinson, D. et al. Land-based plastic leakage into the aquatic environment from municipal solid waste \u0026ndash; Waste flow diagram applied to Tuy Hoa City, Phu Yen, Vietnam. \u003cem\u003eWaste Manage.\u003c/em\u003e \u003cb\u003e186\u003c/b\u003e, 226\u0026ndash;235 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Emmerik, T. et al. Seine Plastic Debris Transport Tenfolded During Increased River Discharge. \u003cem\u003eFront Mar. Sci\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHauk, R. et al. Macroplastic deposition and flushing in the Meuse river following the July 2021 European floods. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e, 124025 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu, J., Zhang, Y., Sui, Y. \u0026amp; Chen, S. Lagrangian study of floating debris transport around the Pearl River Estuary in summer. \u003cem\u003eMar Pollut Bull\u003c/em\u003e \u003cb\u003e211\u003c/b\u003e, (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Emmerik, T. et al. Hydrology as a Driver of Floating River Plastic Transport. \u003cem\u003eEarths Future\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLebreton, L., Egger, M. \u0026amp; Slat, B. A global mass budget for positively buoyant macroplastic debris in the ocean. \u003cem\u003eSci Rep\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiro, M., Mikuś, P. \u0026amp; Wyżga, B. First insight into the macroplastic storage in a mountain river: The role of in-river vegetation cover, wood jams and channel morphology. \u003cem\u003eScience Total Environment\u003c/em\u003e \u003cb\u003e838\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMellink, Y. A. M., van Emmerik, T. H. M. \u0026amp; Mani, T. Wind- and rain-driven macroplastic mobilization and transport on land. \u003cem\u003eSci Rep\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez-Fern\u0026aacute;ndez, D. et al. Floating macrolitter leaked from Europe into the ocean. \u003cem\u003eNat. Sustain.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e, 474\u0026ndash;483 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeone, G. et al. Integrating Bayesian Belief Networks in a toolbox for decision support on plastic clean-up technologies in rivers and estuaries. \u003cem\u003eEnviron. Pollut.\u003c/em\u003e \u003cb\u003e296\u003c/b\u003e, 118721 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVriend, P. et al. Rapid Assessment of Floating Macroplastic Transport in the Rhine. \u003cem\u003eFront Mar. Sci\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGurnell, A. M., Bertoldi, W. \u0026amp; Corenblit, D. Changing river channels: The roles of hydrological processes, plants and pioneer fluvial landforms in humid temperate, mixed load, gravel bed rivers. \u003cem\u003eEarth Sci. Rev.\u003c/em\u003e \u003cb\u003e111\u003c/b\u003e, 129\u0026ndash;141 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheung, P. K., Cheung, L. T. O. \u0026amp; Fok, L. Seasonal variation in the abundance of marine plastic debris in the estuary of a subtropical macro-scale drainage basin in South China. \u003cem\u003eSci. Total Environ.\u003c/em\u003e \u003cb\u003e562\u003c/b\u003e, 658\u0026ndash;665 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIvar do Sul, J. A., Costa, M. F., Silva-Cavalcanti, J. S. \u0026amp; Ara\u0026uacute;jo, M. C. B. Plastic debris retention and exportation by a mangrove forest patch. \u003cem\u003eMar. Pollut Bull.\u003c/em\u003e \u003cb\u003e78\u003c/b\u003e, 252 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeiss, L. \u003cem\u003eet al. The Missing Ocean Plastic Sink: Gone with the Rivers\u003c/em\u003e. (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.science.org\u003c/span\u003e\u003cspan address=\"https://www.science.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Emmerik, T. H. M. The impact of floods on plastic pollution. \u003cem\u003eGlobal Sustainability\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTramoy, R., Gasperi, J., Colasse, L. \u0026amp; Tassin, B. Transfer dynamic of macroplastics in estuaries \u0026mdash; New insights from the Seine estuary: Part 1. Long term dynamic based on date-prints on stranded debris. \u003cem\u003eMar Pollut Bull\u003c/em\u003e \u003cb\u003e152\u003c/b\u003e, (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Sebille, E. et al. The physical oceanography of the transport of floating marine debris. \u003cem\u003eEnvironmental Research Letters\u003c/em\u003e vol. 15 Preprint at (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1088/1748-9326/ab6d7d\u003c/span\u003e\u003cspan address=\"10.1088/1748-9326/ab6d7d\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRyan, P. G. \u0026amp; Perold, V. Limited dispersal of riverine litter onto nearby beaches during rainfall events. \u003cem\u003eEstuar Coast Shelf Sci\u003c/em\u003e \u003cb\u003e251\u003c/b\u003e, (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinto, R. B. et al. Exploring plastic transport dynamics in the Odaw river, Ghana. \u003cem\u003eFront Environ. Sci\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaniya, B. et al. Rainfall erosivity and sediment dynamics in the Himalaya catchment during the Melamchi flood in Nepal. \u003cem\u003eJ. Mt. Sci.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 2993\u0026ndash;3009 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValyrakis, M., Gilja, G., Liu, D. \u0026amp; Latessa, G. Transport of Floating Plastics through the Fluvial Vector: The Impact of Riparian Zones. \u003cem\u003eWater (Switzerland)\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorn, M. P., Junge, L. V., Br\u0026uuml;ll, C., Waldschl\u0026auml;ger, K. \u0026amp; Sch\u0026uuml;ttrumpf, H. Terminal settling and rising velocity prediction of macroplastics: Medical face masks as newly emerged objects of concern. \u003cem\u003eScience Total Environment\u003c/em\u003e \u003cb\u003e908\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLebreton, L. et al. Plastics in Freshwater Bodies. \u003cem\u003ePlast. Ocean.\u003c/em\u003e 199\u0026ndash;225. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/9781119768432.ch7\u003c/span\u003e\u003cspan address=\"10.1002/9781119768432.ch7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTramoy, R., Gasperi, J., Colasse, L. \u0026amp; Tassin, B. Transfer dynamic of macroplastics in estuaries New insights from the Seine estuary: Part 1. Long term dynamic based on date-prints on stranded debris. \u003cem\u003eMar. Pollut Bull.\u003c/em\u003e \u003cb\u003e152\u003c/b\u003e, 110894 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMphefu, T., Nkhonjera, G. \u0026amp; Alowo, R. Evaluation of the probability density function of unpredictable flood events in the Mngeni River basin, South Africa. \u003cem\u003eJ. Water Clim. Change\u003c/em\u003e. \u003cb\u003e16\u003c/b\u003e, 1695\u0026ndash;1714 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLotcheris, R. A. \u003cem\u003eet al. Plastic Does Not Simply Flow into the Sea: River Transport Dynamics Affected by Tides and Floating Plants\u003c/em\u003e. (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ssrn.com/abstract=4449742\u003c/span\u003e\u003cspan address=\"https://ssrn.com/abstract=4449742\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoningh, D. et al. Urban river water level increase through plastic waste accumulation at a rack structure. \u003cem\u003eFront. Earth Sci. (Lausanne)\u003c/em\u003e. \u003cb\u003e8\u003c/b\u003e, 28 (2020).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":true,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Macroplastic Pollution, Rivers, Ocean, Seasonality, Retention zones","lastPublishedDoi":"10.21203/rs.3.rs-9052696/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9052696/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eRivers are major transport routes for land-based ocean plastic pollution. Spatiotemporal, geomorphological and hydrological factors that influence river debris passage are poorly understood. To address such knowledge gaps, we employed GPS drifters (n\u0026thinsp;=\u0026thinsp;66\u0026thinsp;+\u0026thinsp;4 pilot) mimicking macroplastics (\u0026gt;\u0026thinsp;5 cm) across three different rainfall phases in the Umgeni River, South Africa. We analysed the drifter trajectories to investigate seasonal transport and retention to estimate debris emission rates into the Indian Ocean. Observed drifter trajectories showed modest seasonal differences in mean daily transport and retention counts. Our pilot drifters captured the 1:50\u0026ndash;100 year return period flood occurring in April 2022 showing substantial flushing downstream to the Indian Ocean. Mean retention durations showed a notable decrease from the dry phase towards both the peak wet and the wet\u0026ndash;dry transition phases, indicating increased debris mobility during the wet season. Drifters were retained frequently in upstream river sections along meanders and vegetated banks, but the downstream estuarine lagoon emerged as the dominant sink with long-term retention limiting drifter exports into the ocean. Our study presents spatiotemporal insights of macroplastic debris transport to inform river plastic transport modelling and effective debris cleanup and mitigation policy frameworks.\u003c/p\u003e","manuscriptTitle":"Lagoons as Ocean Gatekeepers? Seasonal Transport and Retention Dynamics of Floating Macro-debris in the Umgeni River, South Africa","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-31 13:54:22","doi":"10.21203/rs.3.rs-9052696/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"143019108702980529790469819196135553426","date":"2026-03-31T05:38:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"299266081140167082206706773781345448505","date":"2026-03-16T06:37:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-14T00:17:13+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-12T11:39:13+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-09T10:49:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-09T10:48:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-06T16:38:50+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a0fc709f-e8a7-44c8-8fe4-173035b83a65","owner":[],"postedDate":"March 31st, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":65165215,"name":"Earth and environmental sciences/Climate sciences"},{"id":65165216,"name":"Biological sciences/Ecology"},{"id":65165217,"name":"Earth and environmental sciences/Ecology"},{"id":65165218,"name":"Earth and environmental sciences/Environmental sciences"},{"id":65165219,"name":"Earth and environmental sciences/Hydrology"},{"id":65165220,"name":"Earth and environmental sciences/Ocean sciences"}],"tags":[],"updatedAt":"2026-03-31T13:54:24+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-31 13:54:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9052696","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9052696","identity":"rs-9052696","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}