1748-9326/20/11/114089

Abstract

Microplastics (MPs) (<5 mm) are persistent and ubiquitous pollutants in aquatic environments, with over 80% originating from terrestrial sources. While their presence in the Mediterranean Sea has been documented, studies on MP inputs via local streams remain limited. This study quantified and characterized MP abundances along the Na’aman stream (Israel) and its outlet to the Mediterranean Sea, with samples collected before, during, and after the rainy season from established transects along the streambank and from the nearby beach. The spatial distribution and abundances of MPs were examined in relation to seasonal and hydrological conditions. MPs were sorted by type and counted by dissecting microscope. A representative subset of putative MP particles was examined using Fourier transform infrared spectroscopy. The overall mean (±standard error) MP abundance for the Na’aman stream was 231 ± 67 particles kg−1 sediment dry weight, with significant seasonal and spatial variation. MP accumulation increased during dry months and decreased sharply following flood events—dropping by 43% after a 13 mm rainfall event that triggered stream discharge of 3.25 m3 s−1, highlighting the stream’s role in flushing MPs to the sea. The most common MP types were fragments (61.9%), fibers, and foam, with polypropylene accounting for 56.9% of analyzed particles. MP types and polymer compositions varied by site and season, suggesting diverse sources such as domestic wastewater, agriculture, and marine backflow. These findings confirm that ephemeral stream systems can temporarily trap and subsequently remobilize MPs, acting as critical conduits between land-based sources and the marine environment. Targeted mitigation efforts—such as seasonal cleanup campaigns just before the rainy season—could significantly reduce MP flux to the sea in this location. This is the first documented assessment of MP transport from a freshwater stream into the Mediterranean Sea in Israel and may serve as a baseline for future monitoring in similar Mediterranean-climate regions.

Export citation and abstractBibTeXRIS

Plastics constitute at least 85% of marine litter, representing the most harmful and persistent fraction of oceanic waste. Each year, an estimated 11 million metric tons of plastic enter marine environments—a figure projected to nearly triple by 2040, reaching between 23 and 37 million metric tons annually (United Nations Environment Programme 2021). These plastics degrade slowly (Chamas et al 2020), generally fragmenting into smaller particles known as microplastics (MPs).

Plastic pollution and climate change are often treated as separate issues, yet there are clear connections (Ford et al 2022). Climate change accelerates the frequency of extreme weather events, affecting precipitation and flooding, which have the potential to resuspend and redistribute MPs across terrestrial, freshwater, and marine environments (Zhang et al 2020). The Mediterranean is considered a ‘hotspot’ for climate change as well as plastic and MP pollution (Cózar et al 2015) and is therefore a good setting to study plastic pollution in the context of changing weather patterns (Llorca et al 2020).

Lebreton et al (2017) estimated that over two million tons of plastics are washed from land into the ocean every year and Guerranti et al (2020) anticipates that the export of MPs from rivers to the sea may increase by as much as 50% by 2050. Exceptionally high MP abundances have been reported in the Caorle area on the Adriatic coast (Renzi et al 2018), in the Venice Lagoon (Vianello et al 2013), and elsewhere around the Mediterranean basin (van der Hal et al 2017, Martellini et al 2018).

Studies that have examined MPs in river sediments (Eerkes-Medrano et al 2015, Li et al 2018, Zhang et al 2018) determined that the majority have been found in riverbanks, in bed sediments, and in the riparian zone vegetation (Wendt-Potthoff et al 2020). Variations in water level contribute to plastic accumulation on riverbanks and fluctuations in hydrological conditions can lead to resuspension and redistribution of MPs (Wendt-Potthoff et al 2020). The abundances of MPs in streams and rivers may be linked to hydrological factors, such as water level, river discharge, and flow velocity. Heavy rainfall is thought to increase the flux of terrestrial plastic particles to rivers (Windsor et al 2019) via stormwater systems (Hitchcock 2020). Additional plastics and MPs are released from sewers and wastewater treatment facilities when these are inundated during storms (Hitchcock 2020).

In Israel, as much as 80% of the annual precipitation occurs during the winter (Lelieveld et al 2012), which makes it an ideal season to examine the effect of high rain yields on MP transport. In this study, we report on the abundances of MPs found along an Israeli stream and its outlet to the Mediterranean Sea, highlighting key areas for MP accumulation along its course. The few MP studies previously conducted in Israel have focused on MP and macroplastic abundances in the marine environment (7.7 ± 2.4 MPs m−3; van der Hal et al 2017, Segal and Lubinevsky 2023) and in beach sediments (5.9 MPs kg−1 sand; Pasternak et al 2017, Rubin et al 2022, Segal et al 2022). Therefore, this research may serve as a case study for many regions that are susceptible to high-impact weather, that are drier in the summers and have extreme precipitation events in the winters.

2.1. Study area and sampling procedure

The Na’aman stream is a perennial coastal stream in northern Israel that flows approximately 32 km from its headwaters to its outlet in the Mediterranean Sea south of Akko. The stream’s watershed covers an area of approximately 200 km2 and includes a mix of urban, agricultural, and natural land uses. Land use in the basin is characterized by intensive agriculture (including greenhouses, row crops, and fish farms), industrial zones, wastewater treatment facilities, and several residential communities. The stream’s average annual discharge is estimated at 20–30 million m3, with peak flows occurring during the winter rainy season (Lichter et al 2009). The Na’aman also flows through one of Israel’s last remaining coastal salt marshes, the Ein Afek Nature Reserve, before discharging into the sea. The Na’aman has long been one of Israel’s most polluted stream systems, and the beach south of the outlet is closed to the public as a result. The site was chosen for this study due to its accessibility and lack of routine debris removal, making it suitable for monitoring seasonal MP accumulations along the stream course, outlet, and nearby beach.

This study employed a seasonal monitoring approach used by the Israel Oceanographic and Limnological Research Institute, adopted from UNEP (GESAMP 2019) and MSFD’s protocol for marine debris monitoring (Galgani et al 2013). These protocols were written and implemented in order to estimate the amount, distribution, and composition of plastic particles within a given study area.

Four locations (sites) were selected for sampling to evaluate differences in MP abundances along the Na’aman course, outlet, and nearby beach (figure 1). Three replicates were taken at each site in spring (April–May), fall (October), and winter (December) 2021. In order to examine the effect of seasonal flooding on MP concentrations, a ‘flood event’ threshold was established at 13 mm of rain within a six-hour window. Moore et al (2011) determined that these conditions are conducive to the mobilization of small particles within the watershed. This flood threshold was surpassed on 18 December 2021, just prior to sampling (supplementary material, figures S1 and S2).

Figure 1. The Na’aman study area in northern Israel (a), and the Na’aman watershed (b). Modified after Lichter et al (2009) and indicating locations of wastewater treatment plants (WWTP), urban centers and stream flow gauging stations. Changes in the mouth of the Na’aman stream are illustrated in aerial photographs (panels (c)–(f)), and the sampling site locations are indicated in panel (d). (b) Adapted with permissions from Lichter et al (2009). (a), (c) and (d) Maps data: Google, ©2025 Maxar Technologies.

Download figure:

Standard image High-resolution image

Site X is predominantly characterized by salt marsh vegetation and it was established to evaluate MP inputs from land and to represent the upstream region of the study area. Site A represents the downstream segment of the study area and was selected based on its transitional salt-marsh to foredune features and its proximity to the stream outlet. Sites C and D, characterized by predominantly sandy beach surroundings, were selected in order to evaluate the stream’s influence on MP abundances on the seashore, with C located in the outlet area and D representing an adjacent coastal front with no direct connectivity with the stream. At each of these four sites, a transect was established parallel to the shoreline or riverbank on the lowest flotsam line (after Galgani et al 2013). Because the Na’aman stream is subjected to morphological shifts over time along its course (see figures 1(c)–(f)), the sampling locations were slightly adjusted at each sampling survey, to ensure they represented the different segments of the stream.

2.2. Sample processing

During sample preparation, all tools were washed three times with deionized water prior to use, to minimize contamination. Sediment samples retrieved from the field were transferred to stainless steel laboratory trays, dried at 40 °C for 2–4 d and weighed.

Dried samples were sorted into size fractions (>5000; 4999–1000; and 999–300 μm) using metal sieves. Due to the abundance of large sand particles in samples at sites A and X, a filtered NaCl solution (1.2 g ml−1 density) was used to facilitate the extraction of MPs, especially from the 4999–1000 and 999–300 μm size fractions (after Crichton et al 2017). If the bulk of the sample sediments could readily pass through a 300 μm sieve, samples were not density separated. All samples were placed in glass petri dishes for sorting and counting. Plastic particles were sorted manually, separated and classified by type by means of a Motic SMZ-171 dissecting microscope, fine forceps and dissecting needles, and counted. MP counts were then standardized to the number of particles kg−1 dry weight (DW) sediment. Verification that particles were plastic, e.g. fibers and micro-beads that were too small for FTIR analysis (see below) was performed using the hot-needle test (Beckingham et al 2023).

2.3. FTIR analysis

FTIR measurements of the MP samples were performed using a (IS5-ID1) Thermo Nicolet IS5 FTIR with ID1 Transmission Accessory. Plastic particles were selected for FTIR analysis in order to determine: (a) if they were correctly identified as plastic, and (b) their polymer composition. Due to time and resource constraints, it was not possible to analyze all MPs by FTIR. Subsampling was carried out instead, whereby MPs from each type (fishing line, fragment, film, foam, granule, nurdle) were randomly selected from the total study’s samples for FTIR analysis. This process achieved 200 particle reads, which is larger than the minimum particle number (n = 125) suggested by De Frond et al (2023) to accurately represent polymer diversity from across the pooled dataset. The polymer composition was then corrected according to the type percentage of the total counts. Particles <500 μm and fibers were excluded from this analysis because they were too small to generate a spectral reading on this instrument. A background spectrum was recorded using OMNIC Spectra software. The FTIR spectrum from each sample was then automatically compared against an existing polymer database, ‘Plastic Polymer Kit 1.0,’ from the Hawaii Pacific University’s Center for Marine Debris Research. A hit quality index threshold of >60% was used, consistent with similar studies (Klein et al 2015, Masura et al 2015, Veerasingam et al 2020, Uurasjärvi et al 2021).

2.4. Statistical analysis and modeling framework

For typology data, two statistical approaches were used. A PERMANOVA analysis was used to investigate the differences in MP type assemblage between sites and seasons. The calculation of pairwise distances was based on the euclidean distance method, and the analysis was set to run on a 999 permutation matrix. Statistical difference in the polymer composition was examined with a 2-proportions test based on equality of proportions without continuity correction.

To investigate the spatio-temporal dynamics of MPs along the Na’aman stream, MP abundances were first visualized spatially using a vector layer designed to reflect the stream’s geomorphology for each sampling date. The stream was digitized using Copernicus Sentinel-2A (L2A) satellite imagery (supplementary material, figure S3). Following digitization, MP abundances were interpolated using an inverse-distance weighting spatial function.

To validate the observed empirical spatial patterns, generalized additive mixed models (GAMMs) were computed based on a negative binomial family distribution with a log link function. Prior to this workflow, a multiple imputation technique was implemented to reduce the replication bias caused by missing replicates at site X. This robust method handles missing data effectively by generating predicted values based on the original dataset (Alwateer et al 2024). In this study, a Markov Chain Monte Carlo algorithm was employed (Allison 2003), implemented with a ridge-stabilized linear regression Gibbs sampler via the ‘LaplacesDemon’ package (Hall et al 2020), which provides reliable estimates of parameters and standard errors. MP abundance was then modeled as a function of two main terms: time (number of months), treated as an ordinal variable, and site location, treated as a dimensional variable represented by decimal coordinates to account for spatial autocorrelation. Sites were included as a random effect variable.

GAMM diagnostics evaluated residuals, autocorrelation, collinearity, overdispersion, and zero inflation. Refinements were applied only when assumptions were violated. Following the model diagnostics, the final model was tested for spatial autocorrelation using Moran’s I test.

All analyses were conducted in R v4.4.2 (significance level α = 0.05). Specifically, PERMANOVA was implemented with ‘vegan’, GAMMs with ‘mgcv’, and graphics with ‘mgcViz’. Maps were created in ArcGIS Pro v3.3.2.

3.1. MP abundances

A total of 30 bulk sediment samples were collected from the Na’aman stream for this study, and from these 31,542 MP particles were separated, examined and counted. MPs were found in all samples, regardless of season and site location.

MP abundances varied widely between sampling locations (table 1). The mean abundance of MPs (± standard error, SE) was 231 ± 67 particles kg−1 DW sediment across all sites and seasons. The highest abundance of MPs was found at site A, where an average of 603 ± 221 particles kg−1 DW was recorded across all seasons; 54.49% of all MPs in this study were collected at site A. The average abundance at site X was 159 ± 27 particles kg−1 DW. At site D, on the beach, the mean abundance was 18 ± 8 particles kg−1 DW. The abundances upstream of the Na’aman and just before the outlet (site X and A, respectively) were higher than those found at the site adjacent to the stream outlet (site C) and the nearby beach (site D). The highest abundance of MPs (1148 ± 388 particles kg−1 DW) was at site A during fall, and the lowest was at site D (5 ± 1 particles kg−1 DW) during spring (table 1).

Table 1. Mean microplastic (MP) abundances ± standard error (SE), along with the predicted fits from the generalized additive models and their associated confidence interval (CI) across sites and seasons.

Season	Site	Mean MP abundance (particles kg−1 DW sediment) ± SE	Predicted MP abundance (particles kg−1 DW sediment) + CI
Spring	X	77 ± 39	85 (37–200)
Falla	X	226	478 (206–1110)
Wintera	X	178	172 (69–427)
Spring	A	180 ± 97	233 (94–576)
Fall	A	1148 ± 388	862 (368–2020)
Winter	A	480 ± 458	365 (167–797)
Spring	C	48 ± 14	34 (14–84)
Fall	C	46 ± 5	45 (20–99)
Winter	C	266 ± 205	187 (73–476)
Spring	D	45 ± 1	9 (4–21)
Fall	D	40 ± 18	36 (14–92)
Winter	D	8 ± 2	91 (41–203)

aSE is lacking as only one replicate (n = 1) was available for these dates—the missing samples were lost.

On a seasonal basis, MP abundances averaged 134 ± 43 particles kg−1 DW in Spring 2021; 330 ± 1 particles kg−1 DW during fall 2021 and 253 ± 132 particles kg−1 DW in December 2021.

3.2. Typology

There were significant differences in the proportions of MP types among sites and seasons (PERMANOVA, R2 = 0.66, F = 2.34, p < 0.05). The four most abundant MP types were fragments (62%), followed by foam (12%), film (10%), and granules (7%; figure 2). Several types of MPs were found in some seasons and not in others (figure 2), and in only a few of the sites along the river. Fibers were abundant at all spring sites but less prevalent in other seasons, except at the beach site (D). Likewise, granules were a minor component, and evident only at site X, in all seasons, but were the dominant MP type in winter at site C. Foam and film were present at practically all sites in the three seasons, but these MP types were clearly absent at site D in the spring.

Figure 2. The MP particle types across sites, seasons and seasonal average.

Download figure:

Standard image High-resolution image

Among the particles analyzed by FTIR, 20 were classified as fishing lines (filament), 40 as film, 28 as foam, 50 as fragments, 21 as granules, and 41 as nurdles. The proportions of these deviated significantly (2-Proportions test, χ2 = 383.6, df = 7, p < 0.05). Polypropylene (PP) was the most common polymer, accounting for 47% of all particles analyzed (figure 3), including nurdles (54%), fragments (64%), and films (23%). Polyethylene (PE) was the next most abundant type of polymer. Foam was identified as expanded polystyrene (78%) and polyurethane (18%), whereas films were predominantly of PE nature. Fishing lines and fragments were most generally identified as PP and PE.

Figure 3. The polymers identified among the MP particles examined. Note: ABS = acrylonitrile butadiene styrene, PS = polystyrene, EVA = ethylene-vinyl acetate, PU = polyurethane, PVC = polyvinyl chloride, EPS = expanded polystyrene, PE = polyethylene, and PP = polypropylene.

Download figure:

Standard image High-resolution image

Tar was observed in samples collected at site A in May 2021, both as discrete pellets (which were not included in the MP counts) and as coating layers on the surface of large MP particles. Whereas site A is situated near the mouth of the stream, it is nonetheless not a shoreline or a marine station, and it is noteworthy that tar was not observed there in any of the other seasons or sites in this study.

3.3. Spatio-temporal dynamics

Apparent spatio-temporal dynamics of MP abundance were observed in the Na’aman stream over the course of this study, with site A exhibiting notable seasonal variation (figures 4(a)–(c)). MP abundance varied significantly by time and site (table 2; figures 5(a)–(c)), with no evidence of spatial autocorrelation among sites (Moran’s I test, p > 0.05). MP abundances increased progressively from spring to fall (figures 4(a), (b) and 5(a)), corresponding to the end of the rainy season and the height of the dry season, respectively, after a long period of low stream flow. Model outputs indicated a marked 43% decrease in MP abundance in winter, following the first major rainfall and debris-flushing event on 18 December 2021 (figures 4(b), (c) and 5(a)). This reduction was also reflected in the observed macro-debris that accumulated at site A (figures 4(a)–(c)). Spatially, both empirical (figures 4(a)–(c)) and predicted (figure 5(c)) MP abundances were consistently higher at upstream sites (i.e. X and A) when compared to downstream sites (i.e. C and D), and abundances were substantially higher during early fall (figures 4(b) and 5(b)).

Figure 4. The spatial distribution of empirical MP abundance along the Na’aman stream for spring (a), fall (b) and winter (c), with in situ macro-debris levels depicted in photographs below each corresponding sampling period at site A.

Download figure:

Standard image High-resolution image

Figure 5. The 3D visualization of the smooth terms of the GAMM against the linear predictor (at log-link scale) shown at different angles (a)–(c). Note: for visualization purposes, only latitude is depicted, reflecting the predominantly latitudinal orientation of the Na’aman stream.

Download figure:

Standard image High-resolution image

Table 2. Summary statistics of the GAM models to explore the observed MP spatial patterns. Note: the full model output is provided in supplementary material, table S1.

Parametric coefficients	Estimate ± SE	t	p-value
Intercept	4.668 ± 0.166	28.03	<0.0001
Smooth terms	df(smooth term)	F	p-value
s(Month)	1.879	6.998	<0.05
s(Lat, Long)	4.533	9.231	<0.0001

This study has focused on characterizing typology and seasonal spatio-temporal MP abundances in the Na’aman stream, which discharges into Haifa Bay. To the best of our knowledge, this is the first study in Israel to examine MP discharge into the marine environment via freshwater systems; all other studies have focused on MPs on beaches and in coastal and offshore marine waters. Our findings provide initial insights into seasonal MP inputs and transport dynamics in a Mediterranean-climate stream system, with potential implications for future monitoring under changing hydrological conditions.

The mean MP abundance ± SE recorded in this study, 231 ± 67 particles kg−1 DW is similar to that found in a South African stream (mean = 167 particles kg−1 DW; Heinrich et al 2020), but much lower than in a set of Tunisian streams (2340–6920 particles kg−1 DW; Toumi et al 2019). Our statistical modeling revealed significant spatio-temporal variability in MP distribution across the Na’aman study area. Seasonally, MP concentrations increased from spring to fall, likely due to accumulation during prolonged dry periods, and then decreased in winter following the first major flood event, which flushed MPs downstream. Spatially, the highest concentrations were consistently observed at site A (downstream, near the outlet), and site X (upstream), suggesting localized accumulation zones influenced by stream morphology, vegetation, and hydrological conditions. MP type composition also varied significantly by site and season, underscoring diverse sources and transport mechanisms within the watershed.

The observed decrease in MP abundances in the stream following a flood event demonstrates a ‘flushing’ effect where floating and re-suspensible MPs are washed out to sea. Seasonal flooding remobilizes plastics that have accumulated in the riparian zone or riverbank sediments, as observed by Kumar et al (2021). The Na’aman stream experienced a 43% decrease in MP abundances following the first major flood event of the wet season (discharge of 3.25 m3 s−1), on 18 December 2021, just six days prior to our sampling. This falls near the historical median discharge for winter flooding. Analysis of long-term hydrological records from 1944–2021 reveal an increase in the frequency of winter flood events, while the magnitude of annual peak discharges has remained relatively stable over time (R. Chudinov, personal communication, 6 December 2021). This pattern suggests that winter floods are occurring more frequently in the Na’aman study area, but are not necessarily becoming more extreme. Thus, even moderate floods can substantially reduce MP abundances within streams. Hurley et al (2018) found that seasonal flooding in the UK resulted in a 70% reduction in catchment-wide MP abundances, echoing the finding that even low-magnitude flooding has the ability to transport plastics.

Wendt-Potthoff et al (2020) found that high flow in rivers can remobilize plastics that have accumulated in riverbank sediments and riparian zones, thereby resuspending these into the surface waters. Similarly, MPs >333 μm increased seven-fold in surface waters following a rain event (Moore et al 2011). These observations imply that rivers can shift from being sinks to secondary sources of MPs during times of flooding and runoff (Gallitelli et al 2020) and therefore affect MP abundances in surface waters following rainfall (Lebreton et al 2017).

In the Na’aman stream, the X (upstream) and A (downstream) sites increased in MP abundances from the spring to the fall. Similar observations showing the buildup of MPs during periods of low flow and no rain, were observed in other studies of stream and river systems (Moore et al 2011, Wicaksono et al 2021, Xia et al 2023). This changed abruptly after the first winter rain, as MP abundances decreased at the downstream locations.

One of the outstanding lessons learned in this study is that the MP dynamics vary in the different sections of the stream. The physical features of the stream vary from sites X to C, likely affecting the MP distributions, abundances and types (Graca et al 2017, Zhang 2017). Τhe abundance of MPs was highest overall downstream at site A. This area is part of the Na’aman salt marsh with dense vegetation along the banks (Dorchin and Freidberg 2008), which acts as a temporary trap for debris, including MPs, as the debris moves through the stream system (Zhang 2017).

It is noteworthy that MPs at site A originated both from land and from the sea. The marine influence at this site became evident when tar was found in site A samples, soon after a local offshore oil spill in spring 2021 (García-Sánchez et al 2022, Herut et al 2024). Tar was found in numerous sites along the Mediterranean shores of Israel, clearly related to this oil spill. In addition to the tar, it is likely that floating MPs from the sea were also introduced to the stream at this time, thereby contributing to the plastic debris detected at site A. Van Emmerik and Schwarz (2020) found that estuaries act as accumulation zones due to the flow of water and debris from both land and sea. This conclusion is further supported by reports of high plastic abundances near river and stream outlets (Veerasingam et al 2016, van Emmerik et al 2019, van Emmerik and Schwarz 2020).

The coastal beach sediments south of the stream outlet had the lowest abundances of MPs throughout the study. The average abundance at site D across all seasons was 18 ± 8 particles kg−1 DW. This aligns with another study that found an average abundance of 6 particles kg−1 DW in Israeli beaches (Rubin et al 2022).

One of the most common types of MPs found in the Na’aman stream was fragments, believed to form from the erosion of larger items, such as household products (Eerkes-Medrano et al 2015).

In addition to household plastics, agricultural activities utilize a variety of plastics, potentially adding further sources to streams and rivers. Within the Na’aman watershed, there are multiple agricultural activities utilizing plastics for shading, mulching, and fertilization which may release plastics (Eerkes-Medrano et al 2015, Guerranti et al 2020).

Following the first winter rain, there was a 15% reduction in fragments in the Na’aman study area. This type of reduction has also been observed by Hurley et al (2018), who noted a 40% decrease in MP fragments following strong rains in the Irwell and upper Mersey catchments in England. In our study, the percentage of foam MPs decreased from 18% in the fall to 6% in the winter. In line with these findings, Marsay et al (2023) found that 21% of the marine MPs from the winter sampling were polystyrene, which is most commonly associated with foam. This seems to support the idea that foam is washed from the catchment during flooding, as foam tends to float on or in the water column (Yang et al 2021b). These particles may circulate between beach sediments and sea water but are often eventually removed from the marine environment and redeposited on the beach through wind and wave action (e.g. Windsor et al 2019).

One of the potential MPs sources in the Na’aman watershed is a wastewater treatment plant (WWTP) located upstream of the section that was sampled in this study. A study of the MPs in this particular WWTP showed that the treated water released from the plant had a high proportion of microfibers (Ben-David et al 2020). It is interesting to note that in spring at all sites, there were substantial numbers of fibers in the samples and in both spring and winter there was a large proportion of fibers at site D, on the beach. MP fibers are often associated with textiles and domestic (washing machine) wastes and the abundant fibers counted at site D may be related to another, unrelated source. It is noteworthy that fibers were not common in any of the sites sampled in the fall, after an extended dry summer.

Polypropylene (PP) was the most common MP polymer identified among the particles subjected to spectral analysis in this study, followed by polyethylene. Several studies have noted that PP and PE are the most common polymers found in riverine sediments (Klein et al 2015, Rodrigues et al 2018, Osorio et al 2021, Yang et al 2021b), which may be related to the fact that PE and PP make up about 50% of global plastic production, are commonly used in consumer products, and float (Fiore et al 2022). Polymer-specific density further dictates transportation mechanisms, and consequently, the abundance and diffusion of MPs across environmental gradients (Yang et al 2021a). Low-density polymers such as PP, PE and EPS tend to remain at the water surface, undergoing repeated beaching and re-suspension, whereas high-density polymers are more likely to settle and accumulate in benthic environments (Yang et al 2021a).

Heavier and more frequent winter precipitation is one of the outcomes of climate change in the eastern Mediterranean (Vicente-Serrano et al 2025). This leads to greater runoff, which washes MPs from urbanized areas, agricultural fields, and landfills into rivers, streams, and eventually the sea (Haque and Fan 2023, Parvez et al 2024, Zheng et al 2025). Additionally, intense seasonal rainfall can overwhelm water systems and WWTPs, allowing water containing numerous MPs to enter natural waterways, further compounding the problem (Sun et al 2023).

Understanding MP abundances in ephemeral streams is crucial to addressing these challenges. This underscores the need for systematic, long-term monitoring of stream banks as it allows us to better comprehend their impact on ecosystems and develop effective strategies to mitigate their wide-range harmful effects in a changing climate. Targeted mitigation efforts, such as stream and river cleanup campaigns just before the rainy season could significantly reduce MP flux to the sea.

We wish to acknowledge the financial support provided to AF by the Jewish National Fund Fellows’ Scholarship, the Hatter Research Grant, the University of Haifa Graduate Studies Authority’s Excellence Scholarship, and the Eitan Spitzer Scholarship. Special thanks to the Chemistry Department staff at Israel Oceanographic and Limnological Research Institute, to Professor Ruth Shachak Gross, University of Haifa for the use of her FTIR instrument, and to Ms. Carmel Danino for her kind assistance and support throughout the study.

All data that support the findings of this study are included within the article (and any supplementary files) (Ferry 2023).