Microplastics Interception from Riverine Ecosystems and Translocation to Fish Internal Organs

Abstract

Microplastics (MPs) are widespread in aquatic environments, yet the links between environmental contamination, fish interception, and translocation to internal organs remain poorly understood. This study provides the first integrated assessment of MP contamination in an urban river system, quantifying concentrations in the water column and tracing their interception and organ-specific accumulation in a bioindicator freshwater fish, the planktivorous bleak (Alburnus arborella). MPs were characterized by concentration, shape, and size in six organs: gastrointestinal tract, liver, kidney, muscle, brain, and gonads. Remarkable loads of MPs were detected downstream of a metropolitan area and in all examined organs, with the liver and kidney showing the highest concentrations. Trophic ecology emerged as a key driver of active–indirect MP uptake via contaminated prey. Organ-specific translocation pathways were identified, with long fibers dominating in muscle, fragments in the kidney, and small spheres in the brain, the latter suggesting a higher capacity to cross the hematoencephalic barrier. These results demonstrate that size and shape of MPs, together with feeding ecology, govern MP interception and distribution dynamics, highlighting smaller MPs as a particular risk. The findings fill a critical gap in freshwater MP contamination research, underscore the need for further field-based studies on bioaccumulation and long-term exposure effects, and highlight the importance of incorporating MP morphology and ecological interactions into environmental risk assessments.

Introduction

As generally recognized in waterways worldwide, riverine inputs are the primary pathway for microplastics (MPs) entering marine systems¹. Freshwater systems and coastal marine ecosystems are tightly interconnected along a continuous pathway, linked by hydrological and ecological processes that drive the transport, accumulation, and dispersion of MPs across these ecosystem boundaries².

MPs (1 μm − 5 mm^3,4), as emerging contaminants with widespread occurrence and persistence in aquatic environments, raise global concern. Through accumulation, trophic transfer, and biomagnification across food webs^5–9, MPs can interfere with biological functions^10–13 and ecosystem health, also posing a direct risk for humans¹⁴. In this frame, species trophic ecology and interactions with aquatic matrices are expected to drive MP ingestion and accumulation in specific tissues. MP accumulation has been associated with alterations in functional traits, including behavior, development, and species reproductive performance^15–19.

While MPs in marine ecosystems have been extensively studied, the characteristics of MPs transported by rivers and how they interact with and are intercepted by riverine biota, i.e. fish, remain comparatively underexplored^20,21. Rivers act as conveyor belts of MPs from inland, with concentrations and types varying along the river course depending on the level of urbanization and the surrounding anthropogenic activities, due to inputs such as direct waste discharge, surface runoff, wastewater treatment plant and industrial effluents^22–25. This riverine signature is subsequently transferred downstream to marine ecosystems, influencing both the concentrations and morphologies of MPs accumulated by estuarine and coastal organisms^26,27. Nevertheless, a few studies have focused on the MP transport by rivers, particularly from highly urbanized areas, worldwide^28–32, and in Europe^33–38.

Sampling methodologies differ considerably among studies, varying by site, water compartment, type of MPs (i.e., fibers included or not), and the sampling equipment (i.e., trawl, bulk sampling, in situ filtration, mesh sizes)^21,38–40. This leads to substantial variability in data in terms of reported MP shape, size, and concentration. Moreover, the spatial and temporal distribution of MPs along a river course is further shaped by abiotic and biotic characteristics of the drainage basin, as well as by MP particle features^38,41. These factors can also shape the way fish encounter and intercept MPs. While the ingestion of MPs by marine fish has been widely documented⁴², research on their occurrence in freshwater fish stomachs is comparatively recent^43–45.

In the marine environment, the stomach contents of fish feeding at different water layers have been proposed as a proxy to evaluate the concentration of MPs in different compartments⁴⁶. Marine MP distribution and interception by fish are well-documented, with variations depending on fish trophic ecology and feeding habitat^47–49. Depending on MP size, shape, and density, different MPs in the sea can float, remain suspended in the water column, or settle on the seabed⁵⁰, showing a clear separation among different compartments that influences the dynamics of MP distribution and their subsequent interception by fish species⁴⁶. Moreover, biological processes such as ingestion, biofilm formation, adhesion, and sediment reworking by benthic organisms can modify MP properties and further influence their vertical migration and redistribution among these compartments⁵¹. In contrast to marine systems, such dynamics are far less documented in riverine environments^38,45,52 particularly with respect to how MPs are intercepted by fish with different trophic habits and the mechanisms by which particles can translocate to internal organs. Once intercepted, MPs may be expelled via regurgitation or feces, or translocated to organs, as previously observed in organs such as the liver^53,54. Therefore, gastrointestinal contents alone do not provide a complete picture of contamination, translocation, potential bioaccumulation, and trophic transfer risk, making it necessary to investigate other organs as well^9,55.

Most studies on MP contamination focus on the stomach and, less frequently, a few additional organs^56,57, and parallel detection of MPs in both water and fish remains rare. Despite growing interest in MP contamination, integrated studies addressing simultaneously their presence in river water, interception by fish, and their subsequent translocation to internal organs are still lacking. This represents a relevant gap, as MP interception by biota is closely linked to the surrounding environment. The presence of MPs in the gastrointestinal tract generally reflects active uptake associated with feeding behavior and trophic interactions, whereas passive uptake through adhesion or absorption across mucosal surfaces can allow direct contact between environmentally available MPs and specific tissues through tissue penetration and entry into the bloodstream^58,59. These processes, influenced by MP morphological characteristics, can lead to distinct contamination patterns depending on the interception pathway.

Addressing this gap, this study aims to provide the first integrated assessment of MPs in an urban river system by quantifying and characterizing their concentrations in the water column, evaluating their ingestion, and tracing their translocation to multiple organs (brain, liver, kidney, gonads, muscle, and gastrointestinal tract). This approach enables linking environmental contamination to organ-specific biological accumulation, offering novel insights into MP interception, translocation, and potential accumulation in a freshwater planktivorous fish species.

Materials and methods

Sampling area

The sampling area is represented by the final tract of the Tiber River. This is the largest river in the Italian peninsula in terms of catchment area extension and flow, with an annual average of 240 m³/s, reaching maximum values in winter and minimum in summer. The Tiber River flows through densely populated areas, including the city of Rome, with more than 4.5 million inhabitants, over 80% of whom reside in the metropolitan area. Sampling of the water column and fish specimens was carried out in the same site, downstream of the metropolitan area of the city of Rome (Fig. 1).

Fig. 1.

Study area and sampling site. (a) Drainage basin of the Tiber River in Italy; (b) Sampling site (yellow target) located downstream of the urban area of the city of Rome. Maps were generated using QGIS version 3.14.16 (https://qgis.org/). River basin data were obtained from the Italian Institute for Environmental Protection and Research (ISPRA) and additional cartographic layers from the Geoportale Regione Lazio. Background satellite imagery: © 2025 Google.

Water column sampling

To sample the water column, an in-situ filtration system was specifically fine-tuned for this study. Water column sampling was performed in six replicates during May 2023. Samplings were carried out at a distance of 4 m from the riverbank using an extendable pole into which a silicone tube (ø 5 mm), connected to a self-priming pump, was inserted. The tube was equipped with floats and lead weights to maintain a constant depth of 50 cm. This depth was specifically targeted, as bleak is a column-feeding planktivorous species that primarily forages in the water column. In the selected river section, water depth exceeds 2.5 m⁶⁰, ensuring representative sampling of the upper water column while avoiding potential interference from riverbed resuspension. The water flow was filtered through a cascade set of two stainless steel sieves (5 mm, and 60 μm mesh size, respectively). For each sampling, a total of 500 L of water was filtered, after which a subset of this volume (total volume ≥ 6 L) was microfiltered. Microfiltration was conducted for one hour, and the total filtered volume was quantified to assess particle concentration per liter. During the first filtration step, the particulate retained by the 60 μm sieve was then collected and inserted into a sterile glass bottle using a stainless-steel spatula and a squeeze bottle filled with pre-filtered distilled water. Water filtered through the sieves (containing items smaller than 60 μm) was further microfiltered in-situ on a 0.8 μm mixed cellulose ester filter (MCE, Merck AAWG04700) with a Millipore vacuum pump equipped with a Buchner funnel (XX1004724 Millipore) and quantified using a graduated flask. The filters were then placed in sealed Petri dishes. Bottles and Petri dishes were labeled with the corresponding filtered volume, date, time, and weather conditions, and then transported to the LESA (Laboratory of Experimental Ecology and Aquaculture of the University of Rome Tor Vergata) and kept under a fume hood until processing.

Water samples processing

The solution with MPs (between 5 mm and 60 μm) collected in the glass bottles was first subjected to volume reduction. The content of the bottles was rinsed with filtered bi-distilled water (0.2 μm) and filtered on 0.8 μm MCE filter with the Millipore vacuum pump equipped with a Buchner funnel (XX1514700 Millipore).

From this point onwards, both filters, one containing items of 5mm –60 μm and the other those < 60 μm, were processed following the same protocol. A two-steps organic digestion was performed. For the first 24 h, filters were inserted in 250 mL glass beakers and submerged in 10 mL of 30% hydrogen peroxide (H₂O₂, Carlo Erba, 130 volumes) solution at room temperature (25 °C). After 24 h, the filters were removed from the beaker, rinsed with 2 mL of dH₂O, and placed in their own Petri dish to dry. Then, 1% potassium hydroxide (KOH, Carlo Erba, flakes) was added to the liquid fraction at a 2:5 ratio with respect to H₂O₂, to perform the second organic digestion step. After a further 24 h, the digestion solution was transferred to 250 mL glass cylinders where a zinc chloride saturated salt solution (ZnCl₂, 1.7 kg/L, VWR Chemicals) was added for the density separation step and covered with aluminum foil. A stirring sequence was performed following⁶¹. Samples were then left to decant under a fume hood for 24 h at room temperature. After decantation, the supernatant of each sample was collected and refiltered onto its corresponding filter using the vacuum pump. During each transfer and filtration phase, equipment and glassware walls were thoroughly rinsed with dH₂O to avoid the loss of MPs that eventually remained adhered.

Fish specimens sampling

The bleak (Alburnus arborella), a widely distributed zooplanktivorous species, was selected as a bioindicator of MP contamination in freshwater systems. This small cypriniform inhabits the upper water column, primarily consuming plankton and suspended invertebrates (trophic level 2.8 ± 0.3)⁶². The trophic ecology and morphophysiological traits of A. arborella are consistent with those of some other species of the genus Alburnus⁶³, which are widely distributed and abundant throughout Eurasian freshwater ecosystems⁶⁴. Fish sampling was carried out in the final urban stretch of the Tiber River along a section of approximately 500 m, between 41°50’43.28” N 12°28’2.64” E and 41°48’33.55” N 12°25’5.20” E (Fig. 1). Electrofishing was used to sample a total of 56 bleak specimens in May 2023, within authorized environmental monitoring activities approved by the Regione Lazio Authority. Sampling followed standard national protocols for freshwater fish monitoring. In order to minimize handling stress and animal suffering, specimens were euthanized in situ by anesthetic overdose and then transported in containers with dry ice. All procedures were carried out in accordance with relevant national and European legislation and animal welfare guidelines. Before storage, each fish was weighed (analytical balance Mettler Toledo, XS105 DualRange), sexed, and dissected in order to collect and weigh the organs of interest in this study: the whole gastrointestinal tract (GIT), the whole liver, the kidneys, the gonads (if present) and the brain. Immature gonads were excluded from the gonadal dataset to ensure standardization and avoid variability associated with reproductive stage. The dorsal muscle was also sampled on the left side. Each organ was then weighed, wrapped in aluminum foil, labeled, and stored at -20 °C until processing. For each specimen, standard length, total length, fork length, total weight, and weight of each dissected organ were recorded (Table 1).

Table 1.

Averages of the biometric parameters and dissected organs of bleak specimens.

Parameters of bleak specimens (n = 56)	unit	Mean ± SD
Standard length (cm)	cm	8.08 ± 1.47
Total length (cm)	cm	8.25 ± 2.99
Fork length (cm)	cm	8.08 ± 2.16
Total weight (g)	g	7.26 ± 6.32
GIT weight (mg)	mg	398.11 ± 526.04
Liver weight (mg)	mg	127.68 ± 150.81
Kidney weight (mg)	mg	30.20 ± 34.48
Brain weight (mg)	mg	36.44 ± 31.08
Muscle weight (mg)	mg	874.34 ± 1014.61
Gonads weight (mg) (n = 20)	mg	636.09 ± 370.32

Fish organs processing

The organs were thawed and placed in 250 mL glass beakers. Fish organs underwent the two-step organic digestion with 30% H₂O₂, (1:2 w/v) for 48 h, followed by 10% KOH (3:2 v/v) for 24 h, at 25 °C. Following digestion, samples were directly filtered on 0.22 μm MCE filters (Millipore, GSWP04700). The filters were then placed into Petri dishes, labeled, and kept under a fume hood until the next phase.

MP identification and characterization

Following an optimized protocol⁶⁵, Nile Red (NR) fluorescent dye was used for MP staining: 1mL of NR (Sigma-Aldrich, 10 µg/mL acetone) was added to each filter (both water column and fish organ ones) for 15 min and inserted in the vacuum filtration system over a 0.22 μm polytetrafluoroethylene filter (PTFE, GVS Filter Technology). This step was necessary because the NR dissolved the MCE filter: in this way, any item is harvested on the underlying PTFE one. The staining procedure was followed by the addition of 1 ml of acetone for 10 min to eliminate any residuals MCE filter material.

MP identification and image acquisition were performed using an Axiozoom V.16 fluorescent stereomicroscope (Zeiss, Germany) equipped with a 5 MP CCD camera, ZEN 3.8 DLic software, and a motorized stage for automatic fluorescence photo acquisitions. The following setting were used: intensity 100%, gain 1.0×; saturation 1.50; gamma 0.60. Each MP image was acquired automatically under ultraviolet (UV) (Filter Set 02, G 365, LP 420 nm), blue (Filter Set 38, EX 470/40 nm, BP 525/50 nm), and green (Filter Set 20, EX 546/12, BP 575–640 nm) filters, with exposure times of 26 ms, 2.91 ms and 1.01 ms, respectively. Established guidelines were applied to distinguish MPs from material of animal and plant origin⁶⁶. Water samples containing particles larger than 60 μm were observed at 50x magnification, while 112x magnification was settled for all the other samples (1–60 μm water fraction and fish organs).

Using Image J (Fiji, version 1.53t) software, for each MP item, area and maximum and minimum lengths were measured. These parameters were used to categorize each MP according to size and shape (sphere, fiber, foam, film, and fragment).

Statistical analysis

Before performing the statistical analyses, the data were structured to first compare MP concentrations and characteristics across the three matrices (water, gastrointestinal tract, and fish organs), and subsequently to investigate organ-specific patterns within fish organs. Following this rationale, the Results section is organized according to the same scheme. MP concentrations and related variables are reported as mean/median ± standard deviation (SD). The Shapiro-Wilk and Levene’s tests were then used to assess the normality and homogeneity of data variance, respectively. The Kruskal-Wallis (KW) test was applied to identify statistically significant differences between the MP concentration (items/gram of tissue wet weight or MPs/fish) of multiple groups, since data were not normally distributed. When the Kruskal-Wallis test indicated significant differences, Dunn’s pairwise post hoc test for multiple comparisons among independent groups was conducted to pinpoint specific differences after application of the Bonferroni correction. To evaluate differences in MP occurrence (percentage of individuals with contaminated organs), Fisher’s exact test with Bonferroni correction was applied. Differences in MP characteristic among groups were tested using a one-way PERMANOVA (Permutational Multivariate Analysis of Variance), with fixed factors and 9999 permutations. All the statistical analyses were performed using PAST software (version 4.13⁶⁷). The significance level was set at 0.05 (p < 0.05).

Quality assurance and quality control

The investigation of MPs requires the adoption of protocols to prevent and minimize the possibility of external and airborne contamination⁶⁸. For this reason, all procedures were conducted by avoiding the use of plastic tools (where possible), preferring materials such as steel and glass. Glassware and instruments, before and after each use, were washed and rinsed with dH₂O and dried in an oven at 60 °C. To further reduce contamination, cotton laboratory coats were worn, and work surfaces were cleaned with 95% ethanol before and after each use. Aluminum foil was used to cover objects and solutions during procedures. All solutions used (H₂O₂, KOH, acetone) were pre-filtered with Schleicher & Schuell 5891 filter papers. All procedures were performed under a fume hood. As previously suggested⁴, during sampling and lab procedures, blank filters were kept on the bench during sample processing to assess airborne contamination and were subsequently subjected to the entire analytical protocol, thus also serving as procedural blanks. Additional blank filters were processed independently prior to sample analyses to verify that reagents and laboratory equipment did not introduce external contamination.

Results

Comparison among MPs in water samples, fish gastrointestinal tract and fish organs

Microplastic concentrations

No microplastics were found in any of the control blanks of the study. A total of 909 MPs were found in the collected water sample replicates, with a mean concentration of 305.67 ± 92.37 MPs/m³ (0.31 ± 0.09 MPs/L). When expressed per gram of water (to ensure comparability across matrices), MP concentration corresponded to 3.06*10^− 4 ± 9.24*10^− 5 MPs/g.

Among the 56 bleak specimens analyzed (biometric parameters and organ weights summarized in Table 1), a total of 360 MPs were recorded, occurring in 84% of the gastrointestinal (GIT) tracts. The median concentration was 35.23 ± 39.35 MPs/g, corresponding to 4 ± 6.26 MPs/fish. In fish with no detectable MPs in the GIT, MPs were present in other internal organs, with a median concentration of 3.68 ± 4.79 MPs/g of tissue (4.14 ± 3.09 MPs/fish).

MP concentration was highest in the GIT (Fig. 2a), followed by internal organs (KW, H = 32.61, p < 0.05). MP concentrations in both biological matrices were four to five orders of magnitude higher than those measured in the water samples (Dunn’s post hoc test, p < 0.05).

Fig. 2.

Comparison of MP concentration, shape, and size class composition in water, fish gastrointestinal tract, and fish organs. (a) Box and whisker plot of MP concentrations expressed as number of MPs per gram of tissue; (b) MP shape composition; (c) MP size class composition.

MP shape compositions

Differences in MP shape composition were investigated among the three matrices (Fig. 2b). In water samples, fragments were the most common items found, followed by fibers and foam. Fish GITs were characterized by a higher percentage of fibers, followed by spheres and fragments. A similar pattern was observed in fish organs, although fragments were more abundant than spheres. Significant differences in shape compositions were detected among the three matrices (one-way PERMANOVA test with 9999 permutation, total sum of squares = 15.68, within-group sum of squares = 15.02, F-value = 2.181, p-value < 0.05). Specifically, the percentage of fragments and films was significantly higher in water samples than in fish GIT or fish organs, while the percentage of spheres was significantly lower. Fiber percentage did not differ significantly among matrices, while the percentage of foam was significantly lower in fish organs. Overall, shape composition differed significantly only between water samples and fish GIT (Fig. 2b).

MP size compositions

MP size (maximum length) differed among the water samples, fish GIT, and fish organs (Fig. 2c). The size range of MPs was 2–2700 μm in water samples, 2–1297 μm in fish GIT, and 2–875 μm in fish organs. Mean MP length was highest in water samples, intermediate in fish GIT, and lowest in fish organs (301.72 ± 415.14 μm, 220.32 ± 311.42 μm 84.54 ± 137.79 μm, respectively). Differences in maximum length distributions were highly significant among matrices (KW, H = 163.5, p < 0.001). MP area and minimum length showed the same distribution pattern, with larger MPs in water samples (17,927.04 ± 60,423.62 µm², 125.84 ± 162.01 μm), intermediate values in fish GIT (5,201.30 ± 12,356.14 µm², 84.16 ± 124.05 μm) and smaller MPs in fish organs (1,326.68 ± 2,543.55 µm², 46.97 ± 80.81 μm). Differences were significant for both MP areas (KW, H = 274.2, p < 0.001) and minimum lengths (KW, H = 213.9, p < 0.001):

MP size class composition (Fig. 2c) differed significantly between water samples and both fish GIT and fish organs (one-way PERMANOVA test with 9999 permutation, total sum of squares = 12, within-group sum of squares = 10.07, F-value = 9.587, p-value < 0.001). In particular, the highest abundance of the smaller fraction (KW, H = 18.75, p < 0.001) and the lowest relative abundance of the larger fraction (KW, H = 29.81, p < 0.005) accounted for the differences observed. Specifically, in water samples, the 50–250 μm fraction was significantly more abundant than the smallest fraction (2–50 μm) (KW, H = 8.114, p < 0.05) but not compared to the fraction larger than 250 μm. In fish GIT, the medium fraction was significantly lower (KW, H = 16.26, p < 0.05) than the smallest one but not the largest fraction. In fish organs, the frequency of MP abundance followed a significant decrease (KW, H = 61.34, p < 0.005) with increasing size class.

MP translocation in fish organs

MPs were found in every organ, with varying occurrences, concentrations, shape, and size class compositions. A total of 232 MPs were found among the total organs analyzed, with a mean value of 4.14 ± 3.09 MPs per fish. No microplastics were found in any of the procedural blanks of the study.

MP occurrence and concentration in fish organs

MP occurrence differed among fish organs. The liver showed the highest MP occurrence (57%), followed by the gonads (55%). MPs were detected in 50% of kidney, brain, and muscle samples. Excluding GIT, MPs were simultaneously detected in at least one additional organ in approximately 7% of individuals. MP concentrations, expressed as items per gram of tissue and items per fish, did not differ significantly among organs (Table 2). Mean and median MP concentrations varied across organs, with detectable values reported for liver, kidney, brain, gonads, and muscle. Median MP concentration in the kidney was zero, whereas median values were greater than zero in liver, brain, gonads, and muscle (Table 2).

Table 2.

MP concentration in fish organs in terms of number of items per gram (Items/g) and per fish (Items/fish). The mean ± standard deviation (Mean ± SD), the median (Median), and the minimum-maximum range (Min-max) are reported.

Organ	Items/g			Items/fish
	Mean ± SD	Median	Min-Max	Mean ± SD	Median	Min-Max
Liver	37.93 ± 80.11	2.72	0–333.33	1.33 ± 1.6	1	0–6
Kidney	60.36 ± 104.78	0	0–444.44	1 ± 1.4	0	0–5
Brain	29.74 ± 52.79	3.97	0–333.33	0.70 ± 0.8	0.5	0–3
Muscle	3.29 ± 6.99	0.10	0–35.09	0.82 ± 0.97	0.5	0–4
Gonads	4.19 ± 6.16	0.85	0–20.94	1.67 ± 1.73	1	0–4

When expressed as MPs per fish, mean values were generally lower than concentrations expressed per gram of organ (Fig. 3). Higher mean and median values of MPs per fish where observed in the liver and gonads compared to other organs, which showed consistently low values.

Fig. 3.

MP concentration in fish organs. Box and whisker plot of MP concentration expressed as number of MPs per gram of organ (items/g).

MP shape compositions in fish organs

Differences in MP shape composition were observed among fish organs (Fig. 4a). In the liver, kidney and muscle, fibers were the most prevalent items found, followed by fragments and spheres. Conversely, spheres were the most common shape found in brain samples. In the gonads, fragments were the most common items found whereas fibers were less abundant. Significant differences in shape compositions among the organs emerged (one-way PERMANOVA test with 9999 permutation, total sum of squares = 32.26, within-group sum of squares = 27.94, F-value = 4.00, p-value < 0.0001). Specifically, shape composition in brains differed significantly from that of all other organs, except for gonads. Gonads did not differ significantly from other organs, except for the muscle.

Fig. 4.

MP shape distribution and concentration pattern in fish organs. (a) MP shape composition in liver, kidney, muscle, brain, and gonads; (b) Box and whisker plot of MP shape concentration in fish organs by shape. Uppercase letters indicate significant differences in MP shape concentration patterns among organs. Lowercase letters indicate differences within the same organ, and asterisks highlight differences of the same shape among organs.

Shape specific MP concentrations differed significantly among organs (Fig. 4b), with the exception of gonads when compared with the liver and muscle (p < 0.0001).

Within the same organs (Fig. 4b), no significant differences among shape concentrations were detected in the liver and gonads (KW test, p > 0.05). In the kidney, fiber and fragment concentrations were significantly higher than those of films and foams (KW test, H = 14.88, p < 0.001). In muscle, fiber concentration was the highest, and fragment concentration was significantly higher than film concentration (KW test, H = 28.94, p < 0.001). In the brain, spheres represented the predominant shape (KW test, H = 11.19, p < 0.001). Comparison between organs showed that sphere concentration in the brain was significantly higher than in muscle (KW test, H = 14.03, p < 0.005), while fiber concentration in the kidney was significantly higher than in the brain (KW test, H = 15.01, p < 0.005).

MP size compositions in fish organs

MP size distributions differed among fish organs. The minimum MP length found was 2 μm in all organs, except for the gonads (4 μm). Maximum MP length reached 874.07 μm in the liver, 498.02 μm in the kidney, 617.92 μm in the muscle, 98.73 μm in the brain and 664.78 μm in the gonads (Fig. 5). Mean maximum MP length was significantly higher in the muscle and lower in the brain (KW test, H = 33.73, p < 0.001).

Fig. 5.

Mean maximum and minimum MP length measured in fish organs. Histograms showing the average value of the minimum and maximum MP length for each organ. The dotted line indicates the 50 μm threshold.

MP area distributions differed significantly among organs (KW, H = 37.94, p < 0.0001). MPs found in the brain exhibited significantly smaller mean areas (152.81 ± 251.32 µm²) than those found in the kidney (1314.17 ± 2048.27 µm²), muscle and gonads (1855.39 ± 3036.49 µm²). MPs in the muscle showed the largest mean area (2729.38 ± 3600.58 µm²), differing significantly only from those of the liver (831.8 ± 1877.05 µm²).

Minimum MP length followed a similar distribution pattern among organs (Fig. 5). MPs found in the brain had the smallest mean minimum length, significantly different than those found in the liver, kidney, muscle, and gonads (KW, H = 35.26, p < 0.001). Moreover, MPs in the muscle exhibited significantly greater minimum lengths than those found in the liver.

MP size class composition (small fraction: 2–50 μm, medium fraction: 50–250 μm, large fraction: >250 μm) differed among fish organs (Fig. 6a). The smallest fraction was the most abundant in the brain, liver, and gonads, followed by the medium fraction. In the kidney, the smallest and medium fractions showed comparable percentages, while in the muscle the medium fraction was the most representative, followed by the smallest one. In the brain, MPs belonging to the smallest fraction occurred for 92.86% of the total MPs found.

Fig. 6.

MP size class distribution and concentration patterns in fish organs. (a) MP size class composition in liver, kidney, muscle, brain, and gonads; (b) Box and whisker plot of MP size class concentration in fish organs by size class. Uppercase letters indicate significant differences in MP size concentration patterns among organs. Lowercase letters indicate differences within the same organ, and asterisks highlight differences of the same size class among organs.

Significant differences in MP size class composition were observed among organs (one-way PERMANOVA test with 9999 permutation, total sum of squares = 57.39, within-group sum of squares = 51.84, F-value = 4.76, p-value < 0.001). Size class composition in the brain differed from all other organs, except for gonads, while the liver differed significantly from the muscle and gonads.

MP size class concentrations (Fig. 6b) also differed significantly among organs (one-way PERMANOVA test with 9999 permutation, total sum of squares = 37.68, within-group sum of squares = 29.19, F-value = 7.85, p-value < 0.001). In the liver, MP concentration decreased significantly as the size class increased (KW test, H = 18.32, p < 0.001). In the kidney, concentrations of the small and medium size class were significantly higher than that of the larger size class (KW test, H = 8.33, p < 0.005) while no significant differences among size classes emerged in the muscle. Within individual organs, the smallest size class (2–50 μm) showed the highest concentration in the liver, brain, and gonads (brain: KW test, H = 21.93, p < 0.001; gonads: KW test, H = 8.09, p < 0.05). Comparing the concentrations of the same size class among the organs, it emerged that the 2–50 μm fraction in the muscle was significantly lower than in the liver, kidney, and brain, while concentration in brain differed from that in the gonads (KW test, H = 38.98, p < 0.0001). Concentration of medium fraction in the brain was significantly lower than those in the kidney (KW test, H = 16.7, p < 0.001), while no significant differences among organs were observed for the largest size class.

Discussions

The aim of this study was to understand the partitioning of available MPs within a river water column toward interception by the fish gastrointestinal tract and translocation to different organs. In this context, MPs were, for the first time, simultaneously investigated and characterized in environmental matrices and in the fish gastrointestinal tract, liver, kidney, muscle, brain, and gonads. Water samples were collected from the Tiber River, downstream of the metropolitan area of Rome, at the same location where bleak specimens (Alburnus arborella), a column-feeding planktivorous species, were sampled. The bleak represents a suitable and representative bioindicator for investigating MP interception and chronic exposure in riverine environments, as also adopted in previous studies^69,70.

Building on methodological approaches previously applied in related studies and considering the need to process a large number of samples characterized by complex biological matrices, an extraction and identification protocol was optimized to ensure sample comparability and to minimize potential over- or underestimation of MPs. While the use of H₂O₂ during digestion may cause discoloration of certain particles⁷¹, it ensured high digestion efficiency in organic-rich matrices⁴³. Although Nile Red staining does not allow direct polymer identification, the examination of stained particles under multiple fluorescence filters (UV, blue, and green) maximized the reliable detection of MPs in terms of size and shape^65,72, which represent the primary focus of the present work.

From environmental availability to interception and organ translocation

Bioavailability of microplastics in the river water column

The MP concentration in water samples resulted in 305.67 ± 92.37 MPs/m³ (average ± SD; 0.31 ± 0.09 MPs/L). Generally, MP concentration in rivers is higher than that in the marine water column⁴¹. Heterogeneous data are reported regarding the contribution of densely urbanized areas to the concentration and transport of MPs in riverine bodies and the influence of wastewater treatment plants (WWTPs) on these processes^{18,22,73–75}. The investigated river is highly urbanized, and two large WWTPs serving the Rome Metropolis are located upstream of the sampling area. In this study, the MP concentration recorded was much higher than that reported from other European rivers, generally ranging between 0.15 and 58 items/m^{3 76-78,35,37}, and rarely reaching values ​​over 100 items/m^{3 33,34,38,74}. This difference between literature data and our results could be attributed to differences in sampling methodology²¹. The most frequently used manta or plankton nets, with variable mesh sizes (80–330 μm), partially collect large floating MPs, likely with some loss of the smaller and/or denser items, which often are the most abundant⁷⁹, as observed in the present investigation. On the other hand, when sampling is carried out with a non-discrete system, such as pump filtration³⁹ that allows the collection of a fixed and wider dimensional fraction, reported floating MPs are more concentrated^30,80,81. Accordingly, sampling methodology should be aligned with the trophic ecology of the target species, distinguishing, as far as possible, among benthic-, water-column-, and surface-feeding fish.

Another crucial factor undermining the possibility of data comparison among studies is the selective inclusion of certain MP shape while excluding others³⁴.

In this study, water was pumped from the water column (50 cm below the surface, corresponding to the feeding depth of the target species) at 4 m from the riverbank; the MP size range was established between 1 and 5000 μm and no size or shape exclusion occurred during sampling. Thus, a comprehensive assessment of MPs availability in the water column was obtained. Water samples evidenced fragments as the most frequent MP shape, followed by fibers, and last foam. Despite some differences in MP concentration compared with other studies, our data on dominant MP shapes are consistent with those reported for other European rivers, in which fragments^36,74,76,77 and fibers^33,34,78,80 are the most relevant shapes. Fragments are secondary MPs, resulting from the fragmentation of larger plastic objects occurring in riverine environments. In urbanized areas, the main source of fibers derives from urban wastewater, while foams are mainly generated by the breakdown of packaging materials; spheres are generally primary MPs used as additives in industrial products^75,82,83.

The MP size in the water column ranged between 2 and 2700 μm. Specifically, items ranging from 50 to 250 μm were the most abundant, followed by MPs larger than 250 μm; together, these size classes accounted for more than 78% of the total items. These data are consistent with studies that collected a wider size range^34,78,80,84.

Building on the results of this study, conducted downstream of the metropolitan area, future investigations should also compare upstream and downstream sections and address spatial variability along the river continuum.

Occurrence and interception of microplastics in fish

In this study, 84% of bleak contained at least one MP in GIT. This high occurrence is consistent with the high concentration of MPs in the water column. Studies that have investigated MPs in the GIT of riverine fishes report highly variable percentages, ranging between 15% and 100% of the specimens^43,45,54,85. The GIT median concentration of 4.00 ± 6.26 MPs recorded in this study falls within the ranges reported for riverine species^9,59,86. Despite the limited number of available studies and the strong site-specific variability, comparison with congeneric Alburnus species provides useful context for interpreting MP contamination levels and uptake pathways. In Alburnus alburnus, MP occurrence and abundance ranged from 37.6% (in the Danube and its tributaries)⁸⁷ to 19% (in the Gruža Reservoir, Serbia)⁷⁰. Higher MP burdens have been reported for congeneric Alburnus species from Turkish freshwater systems, including A. sellal from the Upper Tigris River (57%; 1.27 ± 1.3 MPs per individual)⁸⁸, Alburnus spp. (3.1 ± 1.37 MPs per individual)⁸⁹, and A. mossulensis from the Karasu River (6.0 ± 5.6 MPs per individual)⁹⁰, while A. tarichi from the Lake Van and A. chalcoides from the Gharesu River (Iran) exhibited the highest burdens (34 ± 13 and 9.45 ± 6.87 MPs per individual, respectively)^91,92.

Notably, in this study, the GIT median MP concentration was significantly higher than that found in water samples, suggesting that the level of interception can be linked to ecological traits associated with passive and active ingestion. In planktivorous fish, such as the bleak, active ingestion can occur during predation directly on large MPs, within the same prey size class, or, alternatively, indirectly on smaller MPs previously concentrated by prey organisms (Fig. 7). This conclusion is reinforced by evidence from congeneric Alburnus species, where shared trophic ecology and feeding strategies lead to similar feeding-mediated MP interception patterns^70,90–92.

Fig. 7.

Conceptual framework of microplastic uptake and translocation in fish. The diagram illustrates how MP bioavailability in water and trophic ecology shape direct and indirect interception pathways and subsequent organ-specific translocation.

Therefore, analysis of the GIT reflects what the fish directly and indirectly ingested, providing an indication of MP interception as a consequence of its foraging and trophic behavior. However, considering that MPs may also be passively intercepted via mucosal absorption and that ingested particles can be rejected or excreted^58,93, MP concentration in GIT do not fully represent the actual level of contamination within organism. In fact, a lower MP concentration was found in bleak organs than in the GIT.

This finding is consistent with evidence from other studies. Although concentrations in internal organs rarely exceed those observed in the GIT, what clearly emerges is a shift in the composition of MPs accumulating within organs, with shapes and size classes less represented in the GIT and more abundant in internal tissues^56,94,95. These include, for example, longer fibers in muscle tissue or smaller, more regular-shaped particles in organs such as the liver and brain, suggesting the presence of time-integrated selective processes driving translocation to internal organs.

Microplastic uptake, retention and translocation across matrices

In MP translocation dynamics, it is also necessary to consider the time dimension. The gastrointestinal tract of fish acts as a temporary container, with a highly variable retention time. Although difficult to generalize and mostly inferred from laboratory studies, MP retention time reflects the interplay between particle characteristics and gastrointestinal tract complexity. In particular, smaller MPs tend to exhibit longer residence times, while regular shapes (e.g., spheres) are generally evacuated more efficiently than elongated or irregular particles, such as fibers. Concurrently, species with more complex gastric structures tend to exhibit longer and more heterogeneous retention times, whereas species with simpler and more linear digestive tracts, such as the bleak investigated in this study, show faster and more uniform evacuation dynamics^93,96–100. This transient behavior contrasts with the accumulation observed in other organs, which represents a memory of the items ingested or assimilated over time. This discrepancy is also evident when comparing MP shapes and sizes between fish organs and GITs with those available in riverine water. Analyzing shape composition therefore becomes essential to determine whether selective MP interception and translocation occurred depending on particle morphology. In particular, in this study, a statistically significant difference was found in shape composition between water and fish GIT. The results clearly showed that the relative abundances of spheres, fragments, and film were responsible for variations among matrices (Fig. 2b). Fibers are the shapes most commonly found in intestinal contents^43,85. In general, spheres are among the least represented shapes in the water column, whereas they are the second and third most frequent in fish GIT and organs, respectively. Specifically, the spheres found in this study have a small size range (5 μm mean, 10 μm maximum), far below the prey size range of bleak and beyond its visual detection capacity. Therefore, this shape is presumed to be taken up passively during feeding or indirectly through ingestion and magnification of particles previously collected by zooplankton organisms. This hypothesis is consistent with some previous investigations focusing on bottom-up transfer through the food web^{5,17,101–103}. The high concentration of spheres observed in fish organs suggests a biomagnification effect driven by their smooth morphology, facilitating movement across organs, as hypothesized in previous studies⁴¹. The similarity between water samples and fish organ composition may also be explained by additional passive interception pathways besides ingestion, such as adherence or mucosal absorption^81,104–107. This translocation process via the circulatory system can directly connect environmentally bioavailable MPs with organs without involving any preselection in the gastrointestinal tract. This interpretation is further supported by the comparable concentration of fragments in fish organs and water samples, and by the comparable levels of fibers across the three matrices. Presumably, shapes such as film and foam, which were significantly less concentrated in organs, may indicate greater difficulty in adhering to or entering organs. Not only shape but also size is a crucial factor in the interception and translocation of MPs^41,108,109.

To determine whether the interception of MPs is an active (direct or indirect) or passive process, and whether translocation within the organs is size dependent, the maximum length, minimum length, and area of each MP found in the three matrices were examined. The average morphometric values differed significantly among matrices, with the highest mean in water samples, followed by the GIT, and the lowest in fish organs. This progressive decrease in morphometric dimensions may indicate a selection process mediated by the trophic vector (prey) for MP availability in the fish GIT. Subsequently, smaller particles appear to translocate more easily into certain organs, suggesting that overcoming biological barriers involves further size selection.

The partitioning of MPs in distinct size ranges (small: 2–50 μm, medium: 50–250 μm, large: >250 μm) provide additional insight into interception and translocation dynamics. In fish organs, MP abundance decreased significantly with increasing size class, differing from patterns observed in water samples and GIT. In the fish GIT, items ranging from 2 to 50 μm and larger than 250 μm represented 43.08% and 34.55%, respectively, and were not statistically different from that of water. This seems to indicate the likelihood that the larger size class can be actively collected during feeding, whereas the smaller fraction may derive from indirect magnification through prey^102,103,110. Overall, 95% of MPs found in organs were less than 500 μm in size, consistent with previous studies^49,56,94. Specifically, the smallest fraction was the most accumulated in organs, probably due to its small size facilitating entry into the circulatory system and even movement to other organs such as the brain^57,111–114. These findings highlight the key role of bleak trophic ecology in MP interception. As a planktivorous fish, bleak can concentrate and actively select MPs available in the environmental matrix, particularly those comparable in size to their prey (Fig. 7). On the other hand, the bleak could also indirectly accumulate MPs previously concentrated in its prey. Finally, translocation to organs appears to be facilitated by morphometric parameters, particularly small size and regular shapes.

Overall, the integrated analysis of environmental matrix, gastrointestinal tract, and organs revealed that fish effectively concentrate MPs through both direct environmental uptake and indirect trophic interception, leading to biomagnification and to a novel size- and shape-dependent selection during translocation.

Organ-specific MP translocation pathways

MPs were detected across all organ types, showing differences in occurrence, concentration, shape, and size composition (Fig. 7). The average number of MPs per fish was 4.14 ± 3.09, a value that is difficult to compare with literature data since, as far as it is known, the present study is the first to analyze five different organs in addition to the GIT and riverine water. Furthermore, differences in reporting units in the literature (MPs per fish vs. per organ or tissue weight), the limited field investigations on MP translocation, and the heterogeneity of laboratory data complicate direct comparison of results. Indeed, the mechanisms, pathways, and, above all, the kinetics of MP translocation into internal organs remain poorly understood. Several laboratory studies on MP translocation have reported variable results compared with field investigations, probably due to limited exposure times of the experiments (ranging from a minimum of one hour to a maximum of 4 months), the use of concentrations not environmentally relevant, and the use of single, regularly shape virgin plastic particles^41,112,115.

Organ-specific accumulation

In the present study, a comprehensive representation of MP distribution, concentration, and morphometric characteristics across fish organs is provided. Despite less than 7% of fish showing simultaneous contamination (GIT excluded), occurrence was highest in the liver, followed by the gonads, while only half of the other organs were contaminated.

Although no statistically significant differences in MP concentrations (items per g or per fish) were detected among organs, their distribution revealed marked heterogeneity. The kidney and liver exhibited the highest mean concentrations, accompanied by large standard deviations, indicating substantial inter-individual variability. In the kidney, the median MP concentration was zero, suggesting that only a few individuals showed elevated MP levels while most kidneys contained no detectable MPs. These relatively high mean concentrations are comparable to those reported for other river fish¹⁰⁸. Both organs typically accumulate significant amounts of MPs due to their roles in filtering large volumes of blood, maintaining homeostasis, processing waste products, and detoxifying circulating blood^{108,112,113,116}.

Muscle and gonads displayed lower mean MP concentrations, with similar high inter-individual variability. In muscle, both mean and median values were low, indicating that most samples contained few or no MPs, whereas gonads showed a similar distribution, with slightly higher mean values but a median close to zero.

By contrast, the brain exhibited a distinct pattern and emerged as the third most enriched organ. Mean MP concentrations were higher than those observed in muscles and gonads but lower than in the kidney and liver. Notably, the brain showed the highest median value among organs, suggesting that a larger proportion of specimens contained moderate MP levels, alongside variability reflecting the presence of extreme values. This finding further support evidence that, under real scenarios, MPs can penetrate the hematoencephalic barrier, as previously demonstrated in marine species^56,57,117. The presence of MPs in the brain is of particular concern, as it can cause direct mechanical damage or influence metabolic and enzymatic activity^{12,13,17,118–120}. The mechanisms underlying translocation are still not clear; however, several dynamics have been proposed, including junction damage, microglial phagocytosis, and retrograde transport from nerve endings^114,121,122.

Collectively, these findings provide field evidence that MP distribution within fish follows organ-specific accumulation pathways, revealing an internal contamination process linked to physiological functions and suggesting the presence of selective translocation dynamics.

Overall, the widespread presence of MPs across analyzed organs reflects the extent of contamination in aquatic organisms and further emphasizes the importance of understanding the mechanisms governing MP translocation and accumulation.

Finally, occurrence and concentrations in fish organs were considerably higher than those found in the same organs of marine fish^{49,56,57,123,124}, as a probable consequence of exposure to more diluted concentrations in marine environments.

Organ-specific shape distribution

From this analysis, it also emerged that MP translocation and accumulation dynamics are strongly influenced by particle shape and size, which represent crucial factors for adhesion, infiltration, and intra- and paracellular movements^55,112,125. Analysis of the MP shape composition in terms of relative abundance evidenced no significant differences among liver, kidney, and muscle, probably due to the high percentage of fibers found. In contrast, the brain differed significantly from the other organs, except for the gonads, as it was characterized by a high percentage of spheres and lower presence of fibers. The ubiquity of fibers in other organs has also been reported in several studies on marine fish^49,55,56,108, supporting the view that fibers represent one of the most readily translocatable shapes. Evaluation of shape-specific concentration patterns revealed organ-dependent selectivity, with the liver and gonads being the least selective. In the kidney, fiber and fragments were the most concentrated shapes; in muscle, fibers predominated, while spheres were most abundant in the brain, thus demonstrating organ-shape selectivity. The elevated concentration of fibers in muscle tissue may derive from systemic circulation, potentially facilitated by muscle contraction. On the other hand, spheres, due to their general small size, can be more easily absorbed through intestinal villi or respiratory mucosa and subsequently enter the bloodstream^41,55,84.

Organ-specific size distribution

The present findings further identify particle size as a key determinant governing accumulation and translocation processes. The average maximum length of MPs increased significantly from the brain to the other organs and was highest in muscle. A similar trend was observed for mean particle area and minimum length, suggesting that these morphometric parameters contribute to translocation processes. Moreover, fiber thickness can strongly facilitate the infiltration of longer MPs^9,115, as observed in the present study.

Regarding size class composition, the brain differed significantly from all other organs except the gonads. The kidney showed percentages of the smallest and medium fractions comparable with those of the liver, highlighting similarities between these two organs, probably related to their purification roles, which expose them to similar MP types¹⁰⁸. Overall, size concentration patterns exhibited clear organ specific distributions, evidencing that each organ selectively accumulates MPs of varied sizes. What emerges as major forcing is that the concentration of the smallest fraction appeared to penetrate biological barriers more easily, favoring its bioaccumulation. Despite the predominance of this smallest size class, MPs larger than 50–250 μm were also found in all organs. Taken together, these results provide field-based evidence under chronic exposure conditions that MP translocation within fish organs is governed by size- and shape-dependent selective mechanisms, highlighting a process not clearly demonstrated before.

These findings contrast with some short-term micro-nanoplastic exposure laboratory experiments^17,119, but are consistent with previous field studies investigating brain contamination^56,57. Laboratory experiments often suggest that MPs larger than 5 μm are unlikely to cross biological barriers and reach internal organ; however, under natural long-term exposure, particles exceeding 90 μm (this study), 95 μm ⁵⁷, and even 500 μm⁵⁶ have been detected in brain tissue. Therefore, further investigations on wild fish species are needed to better elucidate the mechanisms and kinetics of MP translocation within organs under environmentally realistic exposure scenarios.

Conclusion

This study provides an integrated assessment of MP contamination dynamics in riverine ecosystems, linking interception, translocation, and specific accumulation in fish organs. Remarkable MP concentrations were detected both downstream of an urbanized area and within fish organs, with liver and kidney displaying the highest concentrations. Trophic ecology emerged as a key driver of active-indirect MP interception via prey, while variations in MP shape and size composition highlighted distinct translocation pathways to fish organs. Notably, the brain exhibited a high proportion of small spherical MPs, suggesting their enhanced ability to cross the hematoencephalic barrier. These findings highlight that MP accumulation is shaped by particle characteristics and organ-specific physiology, and the prevalence of smaller MPs underscores the need for field-based studies to better understand the mechanisms underlying MP translocation and accumulation in aquatic organisms. This understanding is crucial for accurate ecological risk assessments and the development of effective environmental health strategies.

Author contributions

G.P. co-conceptualized the study, performed the investigation and formal analysis, developed the methodology, curated the data, and wrote the original draft. C.B. and A.R. conceptualized and supervised the study, reviewed and edited the manuscript.

Funding

Declaration.

This research was supported by the PON Blue Growth program through the project “Interventions to support the advanced, integrated, and sustainable development of aquaculture – INSAIL” (CUP - E86C18001440008) and by the PRIN-PNRR project “Competition and Response in Fish Shoals - COREFISH” (CUP C22PnrrP2022ZPMXH_002).

Data availability

The data that support the findings of this study are available from the corresponding author, G.P., upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.