Tracking rates of macroplastic fragmentation in various environmental compartments is fundamentally important for evaluating the risk of plastic pollution, because it provides direct insights into the amount of secondary microplastics released within these compartments1. Field-based information on the rates of macroplastic fragmentation in different environments is, however, very limited1,2,3 especially for rivers4,5,6,7,8. Recent works have, however, hypothesised that river channels can operate as hot-spots of macroplastic fragmentation because of constant movement of water and sediments in this zone which can favour mechanical interactions of macroplastic with water, sediments, and riverbeds5,8. The intensity of these interactions can be particularly high in the case of mountain river channels, where high-energy water and sediment transport coincide with the presence of numerous physical obstacles such as boulders, bedrock, and large wood within the river channel8. Field experiments exploring this process have not yet been conducted. However, obtaining direct information about the rate of macroplastic fragmentation in mountain rivers is crucial for quantifying the production of secondary microplastics in these environments and evaluation of related risks to their river biodiversity9,10, quality of resources they provide for human populations (e.g., water resources11), and understanding the extent to which they can be transported downstream to lowland rivers and oceans5,8.
Here, we propose a simple field-experiment based methodology for quantifying macroplastic fragmentation rates during its transport in river channels. Our methodology implements mass loss quantification of macroplastic objects, previously utilized in laboratory experiments12, to tagged macroplastic objects transported in river channel (Fig. 1). Using this methodology, we have quantified, for the first time, the mass loss of 1-litre PET bottles occurring during their short-term transport (52-65 days) over distances ranging from 0.37 km to 16.27 km in a mountain river channel in the Polish Carpathians, under low- to medium-flow conditions (Fig. 2). The objective of this work is to present this methodology and report the first insights into macroplastic fragmentation in mountain rivers obtained through its application.
Proposed methodology for quantify riverine macroplastic fragmentation
Our methodology combines mass loss quantification of macroplastic objects, previously utilized in laboratory experiments for determining macroplastic fragmentation12, with macroplastic tracking techniques previously used to quantify the travel distance of tagged macroplastic objects transported in river channel13,14. The proposed workflow consists of four steps: (1) measurement of the masses of virgin macroplastic objects, (2) transport of tagged items in the river, (3) repeated measurements of macroplastic object masses, and (4) calculation of object mass loss over time resulting from their transport. The primary advantage of using mass loss as a proxy for macroplastic fragmentation in rivers, compared to other laboratory techniques previously used for quantifying macroplastic degradation and fragmentation2, is its low cost and minimal need for laboratory analysis. Below, we describe how we applied this four-step procedure to quantify the fragmentation rate of 1-litre PET bottles transported in the Skawa River in the Polish Carpathians (Fig. 2).
Measurement of the masses of virgin macroplastic objects
Measurement of macroplastic mass loss as a proxy of its degradation and fragmentation have primarily been employed in laboratory experiments aimed at determining effects of UV radiation, water movement, and biofilm formation on these processes12. In our experiment, we utilised 177 (n=177) virgin 1-litre bottles made from polyethylene terephthalate (PET). Initially, the mass of each bottle was determined (as the mean of triplicate measurements ) using a precise laboratory balance with an accuracy of 0.001g. Subsequently, the bottles were tagged with numbers drawn on the bottle caps and on the foil tag placed inside them (Fig. 3A). Depending on the planned experiment budget, the size of the rivers, the planned duration of the experiment, and the hypotheses being tested, various tracking techniques can also be considered for future works, including GPS, RFID, radio transmitters, and printed items)13,14.
Transport of tagged items in the river
Field experiment was performed in Skawa river (Polish Carpathians), right-bank tributary of Vistula river (largest river in Poland). Having the total length of 96 km, the river originate at 700 m a.s.l. Its channel width ranges from 5 to 40 metres within the study section. The river has mountainous hydrological regime with little hydrological inertia and therefore a considerable amplitude of flow variability. It is characterised by sudden but short-lasting floods. The total catchment area is 1160 km2 and the average annual flow is 11 m3/s. The riverbed is predominantly composed of gravel and cobbles, with some sections of bedrock present in the middle course of the study section. All bottles were sealed with caps (Fig. 2A) and deployed into the river channel at three locations along the Skawa River in the Polish Carpathians on July 11th, 2022 (Fig. 1A-B). These locations were chosen along the 20 km-long study reach of the river, spanning from Osielec Village (location 1) to the Świnna Poręba Dam Reservoir (as depicted in Fig. 2B). After 52 days (September 1st), 57 days (September 6th), and 65 days (September 14th), the study reaches were surveyed by four persons (two on each river bank), enabling them to collect 43 of the previously deployed tagged bottles (as shown in Fig. 2A-C). The travel distances for each bottle were calculated as the thalweg distance between the point of bottle deployment and the location where the bottle was collected along the study reach (measured using an RTK GPS receiver). Subsequently, the collected bottles were transported to the laboratory for cleaning and to measure their mass loss resulting from mechanical fragmentation during their transport in the river channel.
Repeated measurements of macroplastic object masses
The mass loss of macroplastic objects resulting from their transport in rivers was determined by conducting repeated measurements of the dry macroplastic masses before and after their transport. Before measuring the bottle's mass after their transport in the river, we employed a cleaning procedure similar to that used by Gerritse et al.12( Fig. 3B). Initially, the bottles were cleaned with tap water and detergent, followed by a 12-hour incubation period in 30% H2O2 to eliminate biofilms and other organic matter from their surfaces. Then, bottles were rinsed in distilled water and dried at 45°C for six hours. Before drying, the bottles were opened, and the tagging numbers placed inside them before the experiment were removed. After cleaning, biofilm removal, and drying, each bottle was weighed, and the mass loss for each of them was determined in grams.
We accounted for the possibility of the cleaning procedure itself causing a small-scale mass loss, which could potentially overestimate the final results. To assess this error, we conducted a test cleaning on 24 reference bottles, measuring their masses before and after the procedure. The mean value of bottle mass loss during cleaning, determined from the mass loss of the 24 reference bottles (one bottle was excluded due to contamination during cleaning), was found to be 0.021 g (Table S1).
Subsequently, the mass loss values determined for the bottles transported in rivers (n=43) were corrected using the mean value of bottle mass loss occurring during the cleaning procedure (0.021 g) (Table S2) (1).
macroplastic mass losstransport= (massbefore transport-massafter transport)-mass losscleaning procedure (1)
Utilising the corrected mass loss values for the 43 bottles obtained during the 52-65-day experiment, we calculated the yearly mass loss expressed in grams and as a percentage of the initial bottle mass. Additionally, we determined the rate of bottle surface degradation resulting from the calculated mass losses. For this calculation, we used the density of PET plastic (1.38 g/cm³), and we assumed that bottle fragmentation occurs evenly across their entire external surface (~610 cm²).