Assemblage composition

At the level of material categories, the assemblage was dominated by artificial polymers/plastic, which accounted for 587 items (93.9%) of the total 625 recorded items (Table 1). A much smaller contribution corresponded to artificial polymers/plastic composites (22 items; 3.5%), while metal represented only 13 items (2.1%). Non-plastic materials were nearly negligible, with glass/ceramics contributing 2 items (0.3%) and rubber only 1 item (0.2%). This distribution indicates that the assemblage is almost exclusively composed of synthetic polymers, with non-plastic materials playing a negligible role in its overall composition.

Table 1.

Composition of the litter assemblage by material and use categories, showing a strong dominance of plastics (93.9%), primarily associated with food consumption (543 items), with beverage bottles as the most abundant items, while all other categories and non-plastic materials contribute marginally

Table 1. Long description

The table consists of two columns: Material/use and Total.

* Artificial polymers/plastic: 587 total items.

- Cleaning-related: 14 (includes 1 engine oil bottle and 13 cleaning product containers).

- Food consumption-related: 543 (includes 3 caps/lids, 224 drink bottles > 0.5 L, 304 drink bottles ≤ 0.5 L, 7 hard non-foamed containers, and 5 sachets).

- Medical-related: 1 (medical product bottle).

- Personal hygiene and care-related: 18 (includes 1 cosmetic tube, 16 body care bottles, and 1 sachet).

- Recreation-related: 4 (includes 2 foamed plastic items and 2 toys).

- Smoking-related: 6 (Tobacco/Chimo).

- Vehicle-related: 1 (plastic engine oil bottle between 2.5 and 50 cm).

* Artificial polymers/plastic/composite: 22 total items.

- Food consumption-related: 22 (includes 16 Tetrapack juice, 5 Tetrapack milk, and 1 Tetrapack other).

* Glass/ceramics: 2 total items.

- Food consumption-related: 2 (glass bottles).

* Metal: 13 total items.

- Food consumption-related: 5 (includes 2 metal cans and 3 drink cans).

- Personal hygiene and care-related: 2 (aerosol/spray cans).

- Vehicle-related: 6 (includes 3 drums/barrels, 2 paint tins, and 1 vehicle part/battery).

* Rubber: 1 total item.

- Personal hygiene and care-related: 1 (condom).

* Grand Total: 625 items.

At the use-category level, items associated with food consumption were by far the most abundant, comprising 572 items (91.5%; Table 1). This category included beverage bottles, tetra pack containers, food sachets, caps/lids and rigid food containers. The next most frequent categories were personal hygiene and care-related items (21; 3.4%) and cleaning-related items (14; 2.2%), followed by vehicle-related (7; 1.1%), smoking-related (6; 1.0%), recreation-related (4; 0.6%) and medical-related items (1; 0.2%).

Analysis at the item level showed a very strong concentration in a limited number of dominant categories (Table 1 and Figure 3). Plastic drink bottles ≤0.5 L were the most abundant item (304; 48.6%), followed by plastic drink bottles >0.5 l (224; 35.8%). Together, these two categories represented 528 items, or 84.5% of the total assemblage. Other relatively frequent items were plastic body care and cosmetic bottles/containers (16; 2.6%), tetrapack juice containers (16; 2.6%), plastic bottles and containers of cleaning products (13; 2.1%) and plastic food containers made of hard non-foamed plastic (7; 1.1%). All remaining item types were represented by fewer than 1% each.

Figure 3.

Paired heatmaps of assemblage composition and example pictures of transboundary marine litter. Heatmaps showing counts and percentages for material, use and item categories. The figure (heatmaps and pictures) highlights the dominance of plastics and food-related items, particularly beverage bottles. It reveals strong compositional concentration within a limited number of item types.

Figure 3. Long description

The left section consists of three rows of paired heatmaps. Row A, Material, shows Artificial polymers/plastic as the dominant category with counts exceeding 500 and percentages near 80 percent. Other categories including Metal, Glass/ceramics, and Rubber show negligible presence. Row B, Use, identifies Food consumption related items as the primary category, mirroring the high density of the plastic material. Row C, Item, provides a granular breakdown where Plastic drink bottles less than or equal to 0.5 L and Plastic drink bottles greater than 0.5 L represent the highest counts, reaching approximately 300.

The right section displays five photographs of litter found on beaches, each marked with a national flag. Top row: a Nutri-Milk bottle from Nigeria, an Agua S K bottle from Panama, and a metal beverage can from Nicaragua. Bottom row: a pink Flamingo Beach foam board from Aruba and a rusted metal drum labeled Flammable Liquid from China. The heatmaps use a color scale where dark blue or black indicates high frequency and light yellow indicates low frequency.

This compositional pattern is consistent with a large body of literature indicating that plastics, particularly those associated with food consumption and beverage packaging, dominate marine litter worldwide (Mishra et al., Reference Mishra, Fukey and Wankhar2022; Najahi et al., Reference Najahi, Banni, Nakad, Abboud, Assaf, Operato, Belhassen, Gomes and Hamd2025; Cook et al., Reference Cook, Ionkova, Bhada-Tata, Yadav and van Woerden2026). Several studies across diverse coastal settings have consistently identified plastic materials as the primary component of beach litter, often exceeding 80–90% of total items (Ryan et al., Reference Ryan, Dilley, Ronconi and Connan2019; Vlachogianni and Scoullos, Reference Vlachogianni and Scoullos2023). Within this category, single-use consumer products (especially beverage bottles, caps and food packaging) are repeatedly reported as the most abundant items due to their high production volumes, widespread consumption and inefficient waste management (Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2022; Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024; Garcés-Ordóñez et al., Reference Garcés-Ordóñez, Ergas, Baeza-Álvarez, Honorato-Zimmer, López-Xalín, Vásquez, Canals, de Veer, Aguilera, Arias, Bolaños, Aguilar Fallas, Cuenca, de La Torre León, Díaz Córdoba, Gallardo M de los, Hernández, Marcus, Muñoz-Araya, Pérez-Venegas, Sánchez, Segura Estévez, Valdivia-Chavez, Vélez Tacuri, Xajil-Sabán, Praet, Zambrano, Zárate and Thiel2025). Plastic beverage bottles are recognized as a globally ubiquitous litter type, reflecting their durability, buoyancy and capacity for long-distance transport across marine systems (Savage et al., Reference Savage, Chamberlain, Fellows, Jones, Letessier, Llewellyn, Morritt, Rowcliffe and Koldewey2024; Ryan et al., Reference Ryan, Pichegru and Connan2024, Reference Ryan, Makhado, Cebekhulu, Crawford, Dyer, Masotla, Perold and Connan2026).

ATPL structure

At the highest hierarchical level, the assemblage was entirely composed of the Phylum Litter (625 items), with a dominance of the Class Plastics, which accounted for 609 items (97.4%), while Metal (13 items; 2.1%), Glass (2 items; 0.3%) and Rubber (1 item; 0.2%) were only marginally represented (Figure 4). Similar proportions of plastic items have been reported in coastal and marine environments worldwide such as beaches from Japan (Kusui and Noda, Reference Kusui and Noda2003), the Seychelles (Duhec et al., Reference Duhec, Jeanne, Maximenko and Hafner2015), Sri Lanka (Chang et al., Reference Chang Jang, Ranatunga, Mok, Kim, Hong, Rae Choi and Gunasekara2018) and Bulgaria (Simeonova and Chuturkova, Reference Simeonova and Chuturkova2020), as well as in regions such as the Moroccan Mediterranean coast (Maziane et al., Reference Maziane, Nachite and Anfuso2018), Cocos Islands (Lavers et al., Reference Lavers and Bond2019), Senegal (Castro Tavares et al., Reference Castro Tavares, Moura, Ceesay and Merico2020), Gran Canaria (Rapp et al., Reference Rapp, Herrera, Martinez, Raymond, Santana and Gomez2020) and the southeastern Black Sea (Terzi et al., Reference Terzi, Eruz and Ozseker2020), where plastics typically account for more than 80% of total litter. On a global scale, plastics are consistently identified as the dominant component of marine debris (Derraik, Reference Derraik2002; Jambeck et al., Reference Jambeck, Geyer, Wilcox, Siegler, Perryman, Andrady, Narayan and Law2015; Cook et al., Reference Cook, Ionkova, Bhada-Tata, Yadav and van Woerden2026), reflecting their high production rates, persistence and widespread mismanagement. This convergence indicates that the assemblage is structurally constrained at the upper taxonomic levels, reflecting a highly homogeneous material composition.

Figure 4.

ATPL hierarchical structure (Class and Order). Distribution of litter across ATPL levels, showing strong dominance of Plastics at the Class level and Rigid Plastics at the Order level. The figure shows functional and material simplification of the assemblage. Minor categories contribute marginally to the overall structure.

Figure 4. Long description

A radial diagram with a central origin point branching into four primary Class categories.

At the center, the largest branch is Plastics light blue. It splits into five Order categories along the outer perimeter. Rigid Plastics is the dominant category represented by a large solid blue square at 92.16 percent. Other Plastic orders include Composites at 4.16 percent, Non-Rigid Plastics at 0.48 percent, Film forward slash Label at 0.32 percent, and Foams at 0.32 percent.

The second Class is Metal pink. It branches into three orders. Aerosol Cans at 0.64 percent, Drums forward slash containers at 0.64 percent, and Cans at 0.08 percent.

The third Class is Glass dark blue. It leads to a single order also labeled Glass at 0.32 percent.

The fourth Class is Rubber yellow. It leads to a single order labeled Rubber Items at 0.32 percent.

A legend on the left side identifies the color coding for the four classes. Plastics in light blue, Metal in pink, Glass in dark blue, and Rubber in yellow.

At the Order level, the assemblage exhibited a strong functional concentration within Rigid Plastics, comprising 575 items (92.0%), followed by Composites (26 items; 4.2%) and minor contributions from Cans, Aerosol Cans and other categories, each representing less than 1% (Figure 4). This distribution highlights the predominance of rigid packaging items, which combine high buoyancy, structural resistance and low degradation rates, and are therefore particularly prone to long-distance transport and prolonged environmental residence.

At the Family level, the assemblage revealed a distinctly irregular corporate structure, characterized by a combination of dominant producers and a long tail of low-frequency contributors. The most abundant family was Agencia Sky S.A. (121 items; 19.4%), followed by Tingyi Holding Corp. (50; 8.0%), Nongfu Spring Co., Ltd. (39; 6.2%), PepsiCo Inc. (37; 5.9%) and Petite Bottling Company Inc. (31; 5.0%). Additional relevant contributors included Unicola Panamá (27; 4.3%), OLSU Inversiones (26; 4.2%) and China Resources Beverage (24; 3.8%). Despite this concentration, many families were represented by one or two items, indicating high corporate richness but with a clear dominance gradient. This pattern suggests that the assemblage is shaped by globally distributed beverage production systems, where a limited number of producers apply unequal influence over the overall structure (Figure 5).

Figure 5.

Corporate and brand structure (Family–Genus levels). Visualization of the top 10 main producers (families) and brands (genera) contributing to the assemblage.

Figure 5. Long description

The table contains six columns: Family, Genus, Count, Percentage, Rank, and Country.

* Rank 1: Agencia Sky S. A. (Family) with Agua Sky (Genus). Count 116, 18.56 percent. Country: Panama.

* Rank 2: Tingyi Holding Corp. (Family) with Master Kong (Genus). Count 50, 8 percent. Country: China.

* Rank 3: Nongfu Spring Co., Ltd. (Family) with Nongfu Spring (Genus). Count 39, 6.24 percent. Country: China.

* Rank 4: PepsiCo Inc. (Family) with Gatorade (Genus). Count 37, 5.92 percent. Country: Panama.

* Rank 5: Petite Bottling Company Inc. (Family) with Defruts (Genus). Count 26, 4.16 percent. Country: Panama.

* Rank 6: O L S U Inversiones (Family) with Hydra (Genus). Count 26, 4.16 percent. Country: Panama.

* Rank 7: China Resources Beverage (Family) with C’estbon (Genus). Count 24, 3.84 percent. Country: China.

* Rank 8: Demerara Distillers Limited (Family) with Diamond (Genus). Count 17, 2.72 percent. Country: China.

* Rank 9: Petite Bottling Company Inc. (Family) with Luni (Genus). Count 16, 2.56 percent. Country: Panama.

* Rank 10: Petite Bottling Company Inc. (Family) with U N I (Genus). Count 15, 2.4 percent. Country: Panama.

At the Genus and Species levels, this hierarchical organization became more pronounced through strong brand and product-level convergence (Figure 5). The most abundant genus was Agua Sky (116 items; 18.6%), followed by Master Kong (50; 8.0%), Nongfu Spring (39; 6.2%) and Gatorade (37; 5.9%), with additional contributions from Defruts and Hydra (26 items each; 4.2%). At the species level, the assemblage was dominated by bottled water and beverage products, particularly standardized polyethylene terephthalate (PET) containers in the 500 mL to 1.5 L range.

The results found in this work agree with previous studies that have documented similar dominance structures in marine litter worldwide, particularly for beverage-related products (e.g., Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024; Garces-Ordoñez et al., Reference Garcés-Ordóñez, Ergas, Baeza-Álvarez, Honorato-Zimmer, López-Xalín, Vásquez, Canals, de Veer, Aguilera, Arias, Bolaños, Aguilar Fallas, Cuenca, de La Torre León, Díaz Córdoba, Gallardo M de los, Hernández, Marcus, Muñoz-Araya, Pérez-Venegas, Sánchez, Segura Estévez, Valdivia-Chavez, Vélez Tacuri, Xajil-Sabán, Praet, Zambrano, Zárate and Thiel2025; Ryan et al., Reference Ryan, Makhado, Cebekhulu, Crawford, Dyer, Masotla, Perold and Connan2026). Brand audit approaches and provenance analyses have repeatedly shown that a relatively small number of producers account for a disproportionate fraction of plastic litter found on beaches, often dominated by beverage companies and bottled water brands (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024; Cowger et al., Reference Cowger, Willis, Bullock, Conlon, Emmanuel, Erdle, Eriksen, Farrelly, Hardesty, Kerge, Li, Li, Liebman, Tangri, Thiel, Villarrubia-Gómez, Walker and Wang2024). For example, large-scale assessments of plastic bottles across coastal environments have identified strong contributions from multinational corporations such as The Coca-Cola Company, PepsiCo, Tingyi Holding Corp., Nongfu Spring Co., Ltd. and regional bottling systems, reflecting globalized production and consumption networks (Cowger et al., Reference Cowger, Willis, Bullock, Conlon, Emmanuel, Erdle, Eriksen, Farrelly, Hardesty, Kerge, Li, Li, Liebman, Tangri, Thiel, Villarrubia-Gómez, Walker and Wang2024; Ryan et al., Reference Ryan, Pichegru and Connan2024; da Luz et al., Reference da Luz, de Matos and Malafaia2025). Similarly, studies focusing on stranded bottles and caps have demonstrated that specific brands and manufacturing regions dominate assemblages even in remote locations, indicating both long-distance transport and concentrated production sources (Ryan et al., Reference Ryan, Dilley, Ronconi and Connan2019, Reference Ryan, Pichegru and Connan2024, Reference Ryan, Makhado, Cebekhulu, Crawford, Dyer, Masotla, Perold and Connan2026).

ATPI structure

The assemblage exhibits a highly complex structural organization, as reflected by consistently elevated diversity metrics across multiple ATPI components (Table 2). At the compositional level, species richness reached PSR = 178, accompanied by high brand (PGR = 159) and producer richness (PFR = 145), indicating a system composed of numerous distinct products, brands and corporate entities. Diversity indices further confirm this pattern, with a high Shannon index (H′ = 4.13) and Simpson diversity (1–D = 0.96), demonstrating a highly heterogeneous assemblage with low dominance. This is supported by a relatively high evenness (J′ = 0.80) and low Berger–Parker dominance (BPDI = 0.12), indicating that no single product type applies disproportionate control over the assemblage. Standardized richness indices (Margalef = 27.49; Menhinick = 7.12) confirm that this diversity is intrinsic and not an artifact of sample size. Hill numbers further reinforce this interpretation, showing that diversity is maintained across common (q₁ = 62.13) and dominant (q₂ = 26.47) components of the assemblage.

Table 2.

Anthropocene Taxonomy of Plastics Indices (ATPI) summarizing assemblage structure across compositional, corporate, functional and geo-environmental dimensions

Table 2. Long description

The table is organized into five columns: Family, Index, Value, Notes, and Interpretation.

1. Composition Family:

- P S R: Value 178. Species richness of unique products. Indicates multiple consumption pathways.

- P G R: Value 159. Brand richness. Reflects broad commercial diversity.

- P F R: Value 145. Producer richness. Indicates diffuse corporate responsibility.

- H prime (species): Value 4.1293. Shannon diversity. Confirms a highly heterogeneous assemblage.

- 1 minus D (species): Value 0.9622. Simpson diversity. Indicates low dominance.

- J prime (species): Value 0.7969. Pielou’s evenness. Suggests uniform distribution among product types.

- B P D I (species): Value 0.1248. Berger–Parker dominance. Reinforces structural complexity.

- Margalef (species): Value 27.4941. Richness standardized by N. Confirms diversity is not an artifact of sample size.

- Menhinick (species): Value 7.1200. Richness divided by the square root of N.

- Hill q = 1: Value 62.1343. Exponential Shannon. Indicates diversity is maintained beyond rare items.

- Hill q = 2: Value 26.4669. Inverse Simpson. Indicates moderate concentration among dominant products.

2. Corporate Family:

- B D I: Value 0.1856. Brand Dominance Index. Suggests diffuse distribution.

- C D I: Value 0.1936. Corporate Dominance Index. Indicates no single producer monopolizes the assemblage.

- C Div: Value 3.7207. Producer Shannon diversity. Reflects a complex production landscape.

3. Functional Family:

- F Ric: Value 10. Functional richness. Indicates diverse environmental behaviors.

- P M P I: Value 0.5963. Mobility coefficients. Suggests dominance by plastics capable of long-distance transport.

- F T S: Value 0.5963. Mathematically identical to P M P I.

- P P P: Value 0.8974. Persistence coefficients. Indicates strong potential for long-term environmental retention.

4. Geo-environmental Family:

- R P C I: Value 1.8847. Shannon diversity of origin countries. Indicates significant regional and global connectivity.

Note: Results indicate very high diversity (PSR = 178; H′ = 4.13), low dominance, diffuse corporate responsibility (CDI = 0.19), moderate–high mobility (PMPI ≈ 0.60), high persistence (PPP ≈ 0.90) and significant multi-source connectivity (RPCI = 1.88).

At the corporate level, the assemblage displays a diffuse distribution of responsibility, with low dominance at both brand (BDI = 0.19) and producer levels (CDI = 0.19). The relatively high corporate diversity (CDiv = 3.72) indicates that contributions are distributed across a wide range of producers, rather than being concentrated in a small number of dominant companies. This pattern suggests that plastic pollution in the study area is not controlled by a limited set of corporate actors but instead reflects a fragmented and highly distributed production scenario. Such a structure is consistent with globalized consumption systems and supports the interpretation that responsibility is broadly shared across multiple producers (Cowger et al., Reference Cowger, Willis, Bullock, Conlon, Emmanuel, Erdle, Eriksen, Farrelly, Hardesty, Kerge, Li, Li, Liebman, Tangri, Thiel, Villarrubia-Gómez, Walker and Wang2024), complicating attribution and reinforcing the need for collective regulatory approaches, particularly within Extended Producer Responsibility (EPR) frameworks (Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2019, Reference Williams and Rangel-Buitrago2022).

Functional indices reveal that the assemblage encompasses the full spectrum of environmental behaviors (Table 2). Functional richness (FRic = 10) indicates complete representation of plastic categories with distinct ecological and transport properties. The Mobility Potential Index (PMPI = 0.60) and its equivalent formulation (FTS = 0.60) indicate moderate-to-high dispersal capacity, suggesting that a substantial fraction of items is prone to long-distance transport. In parallel, the high Persistence Potential (PPP = 0.90) indicates strong resistance to degradation and prolonged environmental residence times. The coexistence of high mobility and high persistence defines a system capable of both extensive spatial redistribution and long-term accumulation, reinforcing the dual nature of the assemblage as both transboundary and persistent.

At the geo-environmental level, the assemblage is strongly governed by connectivity processes operating at regional to global scales (Table 2). The RPCI = 1.88 indicates moderate-to-high diversity of origin countries, reflecting multiple external inputs rather than dominance by a single source. In the absence of local contributions, this confirms a fully transboundary assemblage shaped by oceanographic transport, maritime activities and regional circulation dynamics. The study area therefore functions as a convergence zone integrating inputs from adjacent continental margins and transoceanic pathways, pointing to the central role of large-scale connectivity in structuring coastal plastic pollution.

Transboundary source structure

The assemblage exhibited a highly heterogeneous geographic origin, comprising items from more than 45 countries, spanning multiple continents and ocean basins (Figure 6). A strong dominance was observed for a limited number of countries, particularly Panama (326 items, 52%) and China (140 items, 22%), followed by secondary contributors such as Malaysia (17, 3%), Ecuador (16, 3%), Venezuela (12, 2%), USA (11, 2%) and Mexico (10, 1.5%). Additional contributions from countries in Africa (e.g., Senegal, Nigeria and Morocco), Europe (e.g., France, Spain and the Netherlands) and Asia (e.g., UAE, Thailand and Vietnam) indicate a broad and globally distributed origin spectrum (Figure 6). This distribution reveals a clear pattern of regional clustering combined with long-distance inputs, consistent with the multi-source nature of marine litter systems. Such diversity reflects the combined influence of land-based sources, maritime activities and oceanic transport processes, which redistribute plastic debris across spatial scales.

Figure 6.

Geographic origin of litter items. Global distribution of source countries and their relative contributions to the assemblage. The figure shows dominance of Panama and China, alongside contributions from multiple continents. It demonstrates the transboundary and multi-regional nature of the system.

Figure 6. Long description

At the top left is a color gradient legend titled Proportion of items by country, ranging from light yellow at 0.2 percent to dark purple at 52.2 percent. Below this is a world map where countries are shaded according to this scale. Dark purple highlights are visible in Central America (Panama) and East Asia (China). Red and orange shades are distributed across North America (U S A), South America (Brazil, Venezuela, Ecuador), Western Europe (France, Spain), and parts of Africa (Nigeria, Senegal) and Southeast Asia (Malaysia). Light yellow shades appear in Australia and parts of South America and Africa.

In the bottom left quadrant, a horizontal bar chart titled Top 10 countries lists the contributors in descending order:

* Panama: 52.2 percent

* China: 22.4 percent

* Malaysia: 2.7 percent

* Ecuador: 2.6 percent

* Venezuela: 1.9 percent

* U S A: 1.8 percent

* Mexico: 1.6 percent

* Senegal: 1.1 percent

* Brazil: 1.0 percent

* Guatemala: 0.8 percent

* Nigeria: 0.8 percent

* Others: 11.2 percent

The bars for Panama and China are significantly longer than the others, with Panama’s bar being the only one in the darkest teal-blue shade, while the rest are light grey except for the Others category.

The assemblage was strongly dominated by the WEST domain (~55%), followed by a substantial MARITIME contribution (~39%), while the EAST (~5%) and TRANSATLANTIC_EAST (~2%) domains represented minor but consistent inputs (Table 3). The dominance of the WEST domain is primarily driven by countries such as Panama, Mexico, Guatemala and other Central American sources, indicating strong regional connectivity within the Caribbean basin. In contrast, the MARITIME domain reflects a diffuse and geographically unstructured signal, consistent with direct disposal from vessels and offshore activities. The EAST contribution is associated with northern South American countries (i.e., Venezuela), while the TRANSATLANTIC_EAST domain includes identifiable inputs from West African nations such as Senegal and Nigeria, providing clear evidence of inter-basin connectivity.

Table 3.

Distribution of litter items across source domains, showing dominance of regional inputs (WEST = 54.56%), followed by substantial maritime contributions (38.56%) and minor inputs from eastern (4.8%) and transatlantic (2.08%) sources

Table 3. Long description

The table consists of seven columns: Source_Domain, n, percent, Distance class, Distance weight, Default transport, and Default dumping.

* Row 1: Source_Domain WEST; n is 341; percent is 54.56; Distance class is Basin; Distance weight is 2; Default transport is Current_Driven; Default dumping is 0.

* Row 2: Source_Domain EAST; n is 30; percent is 4.8; Distance class is Basin; Distance weight is 2; Default transport is Current_Driven; Default dumping is 0.

* Row 3: Source_Domain TRANSATLANTIC_EAST; n is 13; percent is 2.08; Distance class is Global; Distance weight is 3; Default transport is Current_Driven; Default dumping is 0.

* Row 4: Source_Domain MARITIME; n is 241; percent is 38.56; Distance class is Global; Distance weight is 3; Default transport is Maritime_Direct; Default dumping is 1.

Note: Distance classes and transport modes indicate that most inputs are current-driven, while maritime contributions reflect direct offshore inputs.

These patterns provide multiple lines of evidence for distinct transport processes working simultaneously. The dominance of the WEST domain supports the role of regional current-driven transport, associated with Caribbean circulation systems that redistribute debris across adjacent coastal margins (Rangel-Buitrago, Reference Rangel-Buitrago2025a). The presence of African-origin items confirms transatlantic transport pathways, consistent with equatorial current systems capable of transferring buoyant plastics across ocean basins (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024). At the same time, the high proportion of MARITIME items indicates a significant contribution from direct sea-based inputs, including shipping-related disposal and offshore activities (Saliba et al., Reference Saliba, Frantzi and van Beukering2022). This duality between land-based transport and maritime sources is widely recognized in marine litter systems, where both pathways contribute substantially to coastal accumulation (Shumway and Ward, Reference Shumway and Ward2025).

Transport mechanisms further reinforce this interpretation. Items assigned to the WEST, EAST and TRANSATLANTIC_EAST domains are predominantly associated with current-driven transport, reflecting passive dispersal mediated by surface currents, winds and waves, which can redistribute plastics over regional and transoceanic distances. In contrast, the MARITIME domain represents direct deposition pathways, largely independent of oceanographic transport. Attribution confidence varies among domains, with higher confidence for maritime sources and regionally clustered inputs, and moderate confidence for transatlantic contributions due to degradation effects and partial loss of traceability.

Overall, the transboundary source structure found in this work demonstrates that the assemblage is influenced by a combination of regional connectivity, global-scale transport and maritime activities, confirming its nature as a highly connected, multi-source system. The central Caribbean coast of Colombia functions as a convergence zone, integrating inputs from neighboring regions and distant oceanic pathways.

Source contribution indices

The Transboundary Source Index reached its maximum value (TSI = 1.00), since no item in the assemblage could be attributed to a Colombian source, confirming a fully transboundary assemblage. Applying Eq. (2) to the four source domains (SCI = 0.5456, 0.0480, 0.0208 and 0.3856 for WEST, EAST, TRANSATLANTIC_EAST and MARITIME, respectively) yields a SDI of 0.45. Because SDI is bounded here between 0.25 (four equally contributing domains) and 1 (a single dominant domain), this value indicates a moderate level of concentration, confirming that although the WEST domain is the primary contributor, it does not monopolize the assemblage. Instead, the system presents substantial contributions from other domains, particularly MARITIME inputs. This intermediate dominance supports the interpretation of a structured but non-monopolized system, where regional circulation processes coexist with anthropogenic and transboundary inputs.

The MDS = 241/625 = 0.3856 indicates that approximately 39% of the assemblage met the dumping criteria and was consequently assigned to the MARITIME domain, so that this value coincides with the proportional contribution of that domain (Table 3). This strong signal points to the significant role of sea-based human activities, including shipping and fisheries, in shaping coastal plastic accumulation (Apete et al., Reference Apete, Iacovidou, McPhie and Martin2025; Han et al., Reference Han, Zhang, Liu, Liu, Ma and Wen2026). Unlike current-driven inputs, this component reflects direct deposition mechanisms that bypass oceanographic transport pathways, confirming that the system is not exclusively driven by natural circulation but also strongly influenced by anthropogenic processes at sea.

Applying Eq. (4) to the four domains (WEST and EAST, w_i = 2; TRANSATLANTIC_EAST and MARITIME, w_i = 3) yields TCI = [2 × (341 + 30) + 3 × (13 + 241)] / 625 = 1,504 / 625 = 2.41. This value further demonstrates that the assemblage is governed by multi-scale connectivity processes. This value indicates a moderate-to-high level of connectivity, integrating both regional inputs (WEST and EAST domains) and long-distance contributions (TRANSATLANTIC_EAST and MARITIME domains) weighted by transport distance. The elevated TCI confirms that plastic accumulation in the study area results from the interaction of nearby sources and distant inputs, reflecting the combined influence of basin-scale circulation and global transport pathways.

The Caribbean Sea as a connectivity-driven system

The Caribbean Sea functions as a semi-enclosed but highly connected oceanic basin, where circulation is dominated by a persistent westward inflow of Atlantic waters that are transformed into the Caribbean Current and exported toward the Gulf of Mexico through the Yucatán Channel (Figure 7; Johns et al., Reference Johns, Townsend, Fratantoni and Wilson2002). This inflow originates primarily from equatorial and western Atlantic current systems, including the North Brazil Current and the South Equatorial Current, which transport water masses and associated floating debris from the tropical Atlantic into the Caribbean basin (Luko et al., Reference Luko, da Silveira, Simoes-Sousa, Araujo and Tandon2021; da Luz et al., Reference da Luz, de Matos and Malafaia2025). Once inside, these flows converge into the Caribbean Current, a strong westward jet that acts as the principal mechanism for regional-scale transport and redistribution of floating materials (Centurioni and Niiler, Reference Centurioni and Niiler2003; Richardson, Reference Richardson2005).

Figure 7.

Oceanographic connectivity in the Caribbean basin. Map of major current systems influencing plastic transport, including the Caribbean Current, North Brazil Current and transatlantic pathways. In the same way, the determined domains are presented.

Figure 7. Long description

The main map uses white arrows to illustrate oceanographic connectivity. Starting in the Southeast, the South Equatorial Current flows West along the coast of Brazil, transitioning into the North Brazil Current and then the Guyana Current as it moves Northwest. These currents enter the Caribbean Sea, forming the Caribbean Current which flows Westward. A circular P-C Gyre is located South of the main Caribbean Current near Central America. The flow continues North through the Yucatan Channel into the Gulf of Mexico, where it forms a large loop before exiting East through the Florida Straits. To the Northeast, arrows indicate flow in the Sargasso Sea and transatlantic pathways.

An inset map in the bottom-left titled Domains features a legend and colored arrows. A green arrow labeled WEST originates near the Yucatan Peninsula and curves South toward Central America. A blue arrow labeled EAST and a red arrow labeled T_EAST both originate in the East Caribbean and point West toward the center of the basin. A ship icon labeled MARITIME is positioned in the central Caribbean Sea.

At the basin scale, the Caribbean is not an isolated system but part of a broader Atlantic connectivity network, where transoceanic transport pathways link distant source regions. In particular, evidence from drift simulations and field observations demonstrates that plastic debris originating from the west coast of Africa can be transported across the South Atlantic through the combined action of the Benguela Current, the South Equatorial Current and the North Brazil Current, eventually reaching the northern coast of South America and entering the Caribbean system (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024; da Luz et al., Reference da Luz, de Matos and Malafaia2025; James et al., Reference James, Bezerra, Buhler, Cavalcante, Aguilera, Chen, Gros, Hanke, Lemkau, Nelson, Niles, de Oliveira, Pitchford, Radović, Rodgers, Soares, Socolofsky, Summons, Swarthout, Teixeira, Valentine, White, Yang, Zanardi-Lamardo, Zhang and Reddy2026). These pathways enable transatlantic transport over thousands of kilometers, with travel times on the order of months to a year, confirming that marine plastics can function as long-range dispersal units within global circulation systems. In the present dataset, this pathway is directly evidenced by the West African items assigned to the TRANSATLANTIC_EAST domain (Senegal, Nigeria and Sierra Leone), which could plausibly have reached the study area only through this transatlantic route.

Superimposed on this large-scale flow, mesoscale eddies and regional circulation features further enhance connectivity by modifying transport pathways and generating zones of retention and convergence (Carton and Chao, Reference Carton and Chao1999). Cyclonic and anticyclonic eddies propagate westward across the basin, while regional gyres, such as the Panama–Colombia gyre, create localized recirculation patterns (Beier et al., Reference Beier, Bernal, Ruiz-Ochoa and Barton2017) that favor the accumulation of floating debris along the southwestern Caribbean margin. These processes increase residence times and promote the integration of materials transported from different origins (Rangel-Buitrago and Gracia, Reference Rangel-Buitrago and Gracia2026). These retention processes are consistent with the convergence signal observed here, where items originating from more than 45 countries co-occur within a single stretch of coast.

The central Caribbean coast of Colombia operates as a convergence and accumulation zone, where multiple transport pathways intersect. Regional currents deliver materials from the Caribbean basin, transatlantic currents introduce distant inputs from Africa and maritime activities contribute direct offshore sources. The interaction of these processes results in the concentration of transboundary litter observed in this study. Thus, the Caribbean system is best understood as a connectivity-driven transport network, in which ocean currents act as large-scale vectors that redistribute plastic debris across regional, basin and transoceanic scales, ultimately shaping the structure and composition of coastal assemblages.

Multi-source connectivity: Regional, maritime and transoceanic inputs

The connectivity structure of the assemblage emerges from the interaction of regional circulation, direct maritime inputs and transoceanic transport pathways, which together define a multi-source system operating across spatial scales (Figure 7). The dominance of the WEST domain with items from countries such Costa Rica, Jamaica or Panama (e.g., Families: FIFCO, The Coca-Cola Company and Agencia Sky) reflects the central role of regional Caribbean circulation, where the Caribbean Current redistributes floating materials westward across the basin. This current integrates inputs from adjacent coastal regions and promotes lateral exchange between neighboring countries, explaining the strong regional signal observed in the assemblage.

Superimposed on this regional framework, the MARITIME domain represents a distinct and independent input pathway. Unlike current-driven transport, maritime inputs result from direct disposal or loss at sea, associated with shipping routes and fisheries activities. This mechanism introduces materials from countries such China, Malaysia, Singapore (e.g., Families: Nongfu Spring Co., Ltd., Desa Group, Procter & Gamble) into the system without requiring prolonged transport, generating a spatially diffuse signal that is decoupled from oceanographic pathways. The strength of this component highlights the importance of offshore human activities as a major contributor to coastal plastic accumulation.

In contrast, the EAST domain reflects regional to basin-scale connectivity associated with the northern margin of South America. Inputs from countries such as Venezuela and the Guyana region (e.g., Families: Cosméticos Rolda C.A., Suntory Beverage & Food) are transported toward the Caribbean through the interaction of the North Brazil Current, the Guiana Current and coastal circulation systems, which facilitate the northward and westward redistribution of floating materials along the Atlantic margin (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024; da Luz et al., Reference da Luz, de Matos and Malafaia2025; James et al., Reference James, Bezerra, Buhler, Cavalcante, Aguilera, Chen, Gros, Hanke, Lemkau, Nelson, Niles, de Oliveira, Pitchford, Radović, Rodgers, Soares, Socolofsky, Summons, Swarthout, Teixeira, Valentine, White, Yang, Zanardi-Lamardo, Zhang and Reddy2026). Compared to transoceanic pathways, these inputs originate from relatively shorter distances and are integrated more rapidly into Caribbean circulation, representing an intermediate scale of connectivity between local and global processes.

At a larger spatial scale, the TRANSATLANTIC_EAST domain represents true long-distance transport, linking litter from the west coast of Africa (i.e., Nigeria, Senegal and Sierra Leone) with the Caribbean basin. This connectivity is driven by the combined action of the Benguela Current (Hutchings et al., Reference Hutchings, van der Lingen, Shannon, Crawford, Verheye, Bartholomae, van der Plas, Louw, Kreiner, Ostrowski, Fidel, Barlow, Lamont, Coetzee, Shillington, Veitch, Currie and Monteiro2009), the South Equatorial Current (Luko et al., Reference Luko, da Silveira, Simoes-Sousa, Araujo and Tandon2021) and the North Brazil Current (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024), which form a transoceanic pathway capable of transporting buoyant plastics across the Atlantic Ocean. Such materials can travel thousands of kilometers over timescales of several months, eventually reaching the South American margin and entering the Caribbean system. Once incorporated into regional circulation, these transatlantic inputs are redistributed within the basin, contributing to the assemblage despite their distant origin.

Connectivity inside the central Caribbean of Colombia operates through overlapping transport mechanisms, where regional currents, maritime inputs and transoceanic flows interact to shape the assemblage. The resulting system is not dominated by a single process, but instead reflects a multi-scale connectivity network, in which local accumulation patterns emerge from the integration of diverse and spatially distributed sources.

Functional behavior of plastics in a connected system

The intrinsic physical properties of plastics, particularly their mobility and persistence, determine their role in connectivity-driven systems like those observed here in the central Caribbean of Colombia. The moderate-to-high values of the PMPI (≈ 0.60) indicate that a substantial fraction of the assemblage is composed of items with strong dispersal capacity (i.e., rigid plastics). This mobility is primarily controlled by buoyancy, shape and size, which allow plastics, especially rigid packaging such as bottles, to remain at or near the surface, where they are efficiently transported by currents, winds and waves (van Sebille et al., Reference van Sebille, Aliani, Law, Maximenko, Alsina, Bagaev, Bergmann, Chapron, Chubarenko, Cózar, Delandmeter, Egger, Fox-Kemper, Garaba, Goddijn-Murphy, Hardesty, Hoffman, Isobe, Jongedijk, Kaandorp, Khatmullina, Koelmans, Kukulka, Laufkötter, Lebreton, Lobelle, Maes, Martinez-Vicente, Morales Maqueda, Poulain-Zarcos, Rodríguez, Ryan, Shanks, Shim, Suaria, Thiel, van den Bremer and Wichmann2020; Kaandorp et al., Reference Kaandorp, Lobelle, Kehl, Dijkstra and van Sebille2023; Sanness Salmon et al., Reference Sanness Salmon, Baker, Kozarek and Coletti2023). As a result, plastics function as passive dispersal units, capable of traveling across regional and transoceanic scales.

In parallel, the high Persistence Potential (PPP ≈ 0.90) reflects the resistance of plastics to physical, chemical and biological degradation, enabling prolonged residence times within the marine environment. This persistence allows plastics to remain in circulation long enough to be transported across multiple connectivity pathways, increasing the likelihood of redistribution, re-stranding and re-mobilization (Kane and Clare, Reference Kane and Clare2019; Chen et al. Reference Chen, Li, Bowen and Coco2023). The combination of durability and low degradation rates ensures that plastics are not only transported over long distances but also retained within the system over extended periods (Rangel-Buitrago and Gracia, Reference Rangel-Buitrago and Gracia2026).

The interaction between mobility and persistence generates a dynamic balance between transport and retention processes (Russell et al., Reference Russell, Pohl and Fernández2025). Highly mobile items (i.e., bottles) are redistributed across spatial scales through ocean circulation, while persistent materials accumulate in convergence zones, where hydrodynamic conditions favor deposition. These processes are not independent: mobility enables exposure to multiple transport pathways, while persistence allows repeated cycling between transport and accumulation phases (van Sebille et al., Reference van Sebille, Aliani, Law, Maximenko, Alsina, Bagaev, Bergmann, Chapron, Chubarenko, Cózar, Delandmeter, Egger, Fox-Kemper, Garaba, Goddijn-Murphy, Hardesty, Hoffman, Isobe, Jongedijk, Kaandorp, Khatmullina, Koelmans, Kukulka, Laufkötter, Lebreton, Lobelle, Maes, Martinez-Vicente, Morales Maqueda, Poulain-Zarcos, Rodríguez, Ryan, Shanks, Shim, Suaria, Thiel, van den Bremer and Wichmann2020; Rangel-Buitrago, Reference Rangel-Buitrago2025a).

Plastics can be conceptualized as mobile persistent particles, whose environmental behavior is analogous to that of long-lived tracers within ocean circulation systems (Russell et al., Reference Russell, Pohl and Fernández2025; Rangel-Buitrago et al., Reference Rangel-Buitrago, Ben-Haddad, Nicoll, Galgani and Neal2026a). This dual property explains their capacity to connect distant regions, integrate multiple sources and sustain accumulation patterns over time. Consequently, the structure of marine litter assemblages is not only a function of source inputs but also of the intrinsic properties of plastics, which govern their transport trajectories and retention within the marine system.

Corporate structure and distributed responsibility

The corporate structure of the assemblage reveals a system characterized by distributed rather than concentrated responsibility, where no single producer applies full control over the observed litter composition. Although certain companies contribute disproportionately to the assemblage, the overall pattern is defined by many producers/families (145) with varying levels of representation. This is consistent with the low dominance values observed in the ATPI metrics, indicating that the system is not monopolized but instead reflects a fragmented corporate scenario.

At the producer (Family) level, a limited number of companies and countries emerge as dominant contributors, including Agencia Sky S.A. (Panama), Tingyi (Cayman Islands) Holding Corp. (China), Nongfu Spring Co (China)., Ltd., PepsiCo Inc. (Panama) and Petite Bottling Company Inc. (Panama), among others (Figure 8). These producers, primarily associated with bottled beverages and consumer packaging, reflect the global scale of plastic production and distribution systems. Despite their prominence, they coexist with a long tail of less frequent contributors, including regional bottlers and smaller companies, which collectively contribute to the overall diversity of the assemblage. This pattern establishes that marine litter is not the result of isolated actors, but rather the outcome of interconnected and multi-scalar production networks.

Figure 8.

Corporate connectivity network. Network-style representation linking major producers and brands to the central Caribbean system. The figure highlights key corporate contributors and their spatial associations.

Figure 8. Long description

At the center of the diagram is an oval labeled Central Caribbean of Colombia. Eight blue lines radiate outward to corporate logos and associated photographs of discarded plastic bottles.

* Top-Left: PEPSICO logo with two photos of orange and green labeled bottles.

* Top-Center: TINGYI HOLDING CORP. logo with four photos of various tea and water bottles.

* Top-Right: AguaSky AGENCIA SKY logo with three photos of clear water bottles.

* Right: Wahaha logo with two photos of small clear bottles with red labels.

* Bottom-Right: C R BEVERAGE logo with two photos of blue-labeled water bottles.

* Bottom-Center: THE Coca-Cola COMPANY logo with five photos including Sprite, Coca-Cola, and various bottle caps.

* Bottom-Left: Petite Bottling Company logo with four photos of clear water bottles.

* Left: NONGFU SPRING logo with one photo of a large water bottle with a red label.

Each corporate section displays hands holding the plastic debris against a background of sandy beach and coastal vegetation.

At the brand (Genus) level, similar patterns of partial concentration are observed, with dominant brands such as Agua Sky (Panama), Master Kong (China), Nongfu Spring (China) and Gatorade (Panama) contributing significantly to the assemblage (Figure 8). These brands are widely distributed across global markets and are strongly associated with single-use plastic packaging, particularly PET bottles (Garces-Ordoñez et al., Reference Garcés-Ordóñez, Ergas, Baeza-Álvarez, Honorato-Zimmer, López-Xalín, Vásquez, Canals, de Veer, Aguilera, Arias, Bolaños, Aguilar Fallas, Cuenca, de La Torre León, Díaz Córdoba, Gallardo M de los, Hernández, Marcus, Muñoz-Araya, Pérez-Venegas, Sánchez, Segura Estévez, Valdivia-Chavez, Vélez Tacuri, Xajil-Sabán, Praet, Zambrano, Zárate and Thiel2025; Ryan et al., Reference Ryan, Makhado, Cebekhulu, Crawford, Dyer, Masotla, Perold and Connan2026). Their presence across distant regions points to the role of globalized consumption systems, where standardized products are produced, distributed and consumed across multiple countries (Tian et al., Reference Tian, Zhong, Chen, Feng, Sun, Zhang, Shao, Liu and Hubacek2024), increasing the likelihood of transboundary dispersion once they enter the marine environment.

The coexistence of dominant producers with a high diversity of minor contributors results in a system of diffuse responsibility, where environmental impacts cannot be attributed to a single company or region. This poses a major challenge for governance, as traditional regulatory approaches often focus on local or national sources, while the observed assemblage reflects a globally distributed production–consumption network (Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2022). The fragmentation of responsibility complicates accountability and highlights the limitations of approaches that do not consider the transboundary nature of marine litter (Shi and Zhang, Reference Shi and Zhang2026).

These findings have direct implications for EPR frameworks. The absence of a single dominant producer and the presence of multiple contributing companies suggest that effective mitigation strategies must move beyond isolated corporate actions toward collective and coordinated responsibility mechanisms. Policies should target high-frequency product categories, particularly single-use beverage packaging and promote standardized solutions such as reusable systems and cross-border return schemes (Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2019). At the same time, the global distribution of producers makes clear the need for international governance frameworks, capable of addressing plastic pollution across the full production–consumption–disposal chain. In this context, marine litter should be understood not only as an environmental issue but also as a systemic outcome of global economic structures, requiring integrated policy responses at regional and global scales (Bellou et al., Reference Bellou, Gambardella, Karantzalos, Monteiro, Canning-Clode, Kemna, Arrieta-Giron and Lemmen2021; Rangel-Buitrago, Reference Rangel-Buitrago2025a; Shumway and Ward, Reference Shumway and Ward2025).

Conceptual model: Transboundary plastic connectivity

The information presented in this work allows the development of a conceptual model describing how transboundary plastic pollution operates as a connectivity-driven system in the central Caribbean coast of Colombia (Figure 9). This model is based on the empirical integration of source attribution, hierarchical organization and connectivity processes, derived from the combined application of the ATPL, the ATPI and the quantitative characterization of transboundary domains. This model synthesizes the observed relationships between sources, transport pathways and assemblage structure, conceptualizing marine plastic pollution as an organized anthropogenic system functioning across multiple spatial scales. It is structured around four primary components: source domains (transport mechanisms, environmental convergence and assemblage organization), which together define the spatial configuration, internal structure and dynamic behavior of plastic litter in the marine environment.

Figure 9.

Conceptual model of transboundary plastic connectivity. Integrated model linking source domains, transport pathways, convergence zones and assemblage organization. The figure synthesizes empirical results into a system-level framework. It conceptualizes marine litter as a connectivity-driven anthropogenic assemblage.

Figure 9. Long description

The flowchart begins at the top with a central box labeled SOURCE DOMAINS. Four arrows point downward from this box to four sub-categories: West Regional Inputs, East N S A Margin, Maritime Offshore Direct, and Transatlantic_East Long Distance Input.

Arrows from these four sub-categories converge into a single box labeled Transport Pathways. This box points down to a bulleted list of specific mechanisms:

• Caribbean Current regional redistribution

• North Brazil Current basin-scale input

• South Equatorial Current transatlantic

• Maritime direct inputs shipping or fishing

A single arrow points from this list down to a box labeled Convergence / Sink Zone Central Caribbean Coast of Colombia.

Below this, an arrow leads to Assemblage Organization, which then points to a final central box titled Connectivity-driven Anthropogenic Assemblage.

From this final box, three arrows point downward to the concluding descriptive categories:

1. Material Composition Plastics Dominance

2. Corporate Distribution Multi-Producer

3. Functional Traits Mobility, Persistence

At the source level, the model identifies four principal domains: WEST, EAST, TRANSATLANTIC_EAST and MARITIME. These four domains represent distinct geographic and functional origins of plastic inputs, ranging from regional land-based sources to transoceanic contributions and direct offshore disposal. The WEST domain constitutes the dominant regional input, reflecting the influence of Caribbean circulation and adjacent coastal systems. The EAST domain represents contributions from the northern South American margin, while the TRANSATLANTIC_EAST domain captures long-distance inputs originating from West Africa. The MARITIME domain operates independently of land-based pathways, representing direct inputs from shipping and offshore activities.

Transport processes form the central dynamic component of the model. These include regional circulation, primarily driven by the Caribbean Current (Richardson, Reference Richardson2005), as well as transoceanic pathways associated with the South Equatorial and North Brazil currents (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024), which enable the movement of plastics across the Atlantic basin. In parallel, maritime activities introduce direct inputs that bypass oceanographic transport. The interaction of these mechanisms generates a multi-scale transport network, where materials are redistributed across regional, basin and global scales.

The convergence of these transport pathways results in the formation of accumulation zones, where hydrodynamic conditions favor the retention and concentration of floating debris. The central Caribbean coast of Colombia is conceptualized within this model as a sink or convergence zone, where inputs from multiple domains intersect (Ben-Haddad et al., Reference Ben-Haddad, Abelouah, Hajji, Rangel-Buitrago and Alla2023, Reference Ben-Haddad, Abelouah, Hajji, Abou Oualid, Rangel-Buitrago and Ait Alla2024). Environmental features such as coastal geomorphology, circulation patterns and local retention processes act as filters that enhance accumulation and prolong residence times (Rangel-Buitrago et al., Reference Rangel-Buitrago, González-Fernández, Defeo, Neal and Galgani2024).

At the system level, the assemblage is represented as a hierarchically organized structure, where material composition, corporate distribution and functional traits interact to define its configuration. High diversity and distributed contributions reflect the integration of multiple production and consumption systems, while the persistence and mobility of plastics enable continuous redistribution within the system. As a result, the assemblage is not a static accumulation, but a dynamic entity shaped by ongoing inputs and transport processes.

This conceptual model illustrates that marine plastic pollution is governed by interconnected processes operating across spatial scales, linking distant sources to localized accumulation zones. In the same way, this model serves as a basis for understanding plastic pollution as a global, system-level phenomenon, with implications for both scientific analysis and environmental management.

Environmental and socioeconomic implications of transboundary marine litter

The connectivity-driven nature of the assemblage found has significant implications for coastal ecosystems, where the continuous influx of transboundary plastics alters both physical and ecological processes. Accumulated debris can modify sediment dynamics, interfere with habitat structure and reduce environmental quality (Browne et al., Reference Browne, Underwood, Chapman, Williams, Thompson and van Franeker2015). In beach environments, large quantities of plastic items (particularly bottles and rigid packaging) can influence sediment retention and surface roughness (Harris et al., Reference Harris, Westerveld, Nyberg, Maes, Macmillan-Lawler and Appelquist2021), while in adjacent habitats such as mangroves and seagrass beds, plastics can become entangled within vegetation, affecting growth and stability (Rangel-Buitrago et al., Reference Rangel-Buitrago, González-Fernández, Defeo, Neal and Galgani2024). These impacts are not limited to local inputs but are amplified by the continuous arrival of externally sourced materials, reinforcing the role of connectivity in sustaining pollution levels.

From an ecological perspective, plastic debris acts as both a physical and biological stressor. Marine organisms are affected through ingestion, entanglement and habitat disruption, with documented impacts across trophic levels (Schmidt et al., Reference Schmidt, Kühnel, Materić, Stubenrauch, Schubert, Luo, Wendt-Potthoff and Jahnke2024). In addition, plastics serve as vectors for contaminants and microorganisms, facilitating the transport of potentially harmful substances and invasive species across large distances (Gracia et al., Reference Gracia and Rangel-Buitrago2018, Reference Gracia, Durán-Fuentes, Santodomingo and Rangel-Buitrago2021). The persistence of plastics enhances their capacity to adsorb hydrophobic pollutants, which can then be transferred through food webs, extending the ecological consequences of plastic pollution well beyond the physical presence of the debris itself.

The implications for tourism and coastal economies are equally significant. Coastal areas such as the central Caribbean coast of Colombia rely heavily on their aesthetic and recreational value, which is directly compromised by visible litter accumulation. The presence of transboundary debris reduces scenic quality, affects visitor perception and can lead to economic losses associated with decreased tourism demand and increased cleaning costs (Rangel-Buitrago, Reference Rangel-Buitrago2019; Rangel-Buitrago and Ben-Haddad, Reference Rangel-Buitrago and Ben-Haddad2024). Importantly, the external origin of much of this debris implies that local communities take on the economic load of pollution generated elsewhere, highlighting an asymmetry between source and impact that complicates management and governance.

The study also reinforces the importance of natural traps and accumulation environments in modulating the spatial distribution of plastics. Coastal features such as driftwood accumulations, vegetation zones, embayments and low-energy sectors act as retention structures that enhance the deposition and persistence of debris (Ben-Haddad et al., Reference Ben-Haddad, Abelouah, Hajji, Rangel-Buitrago and Alla2023; Rangel-Buitrago et al., Reference Rangel-Buitrago, González-Fernández, Defeo, Neal and Galgani2024). These environments function as intermediate sinks, where plastics can accumulate temporarily or permanently depending on hydrodynamic conditions. The interaction between connectivity processes and local retention mechanisms therefore plays a critical role in determining where plastics ultimately accumulate within the coastal system.

Limitations and uncertainties

The results presented above are subject to a number of limitations that should be considered when interpreting the source-domain structure and the connectivity metrics derived from it. The first concerns the sampling frame: the dataset is based on six-monthly surveys carried out at three exposed sites between September 2025 and February 2026, and therefore captures seasonal rather than interannual variability. In addition, source attribution requires traceable markings, so the assemblage analyzed here is inherently biased toward labeled, rigid packaging and under-represents fragmented, unbranded or heavily degraded items, which may follow different transport and retention pathways.

The source-domain classification rests on the country of manufacture as a proxy for the source region, and this proxy is necessarily imperfect. Products are frequently manufactured in one country, bottled or distributed in another, and consumed in a third, so the information printed on a label indicates where an item was produced rather than where it entered the ocean. The limitation is most relevant for globally distributed brands and for items manufactured in Asia but circulated through Caribbean or maritime supply chains. Items with partial or degraded labeling were assigned MEDIUM or LOW confidence, and the more conservative attribution was retained wherever forensic and oceanographic evidence diverged; nevertheless, a fraction of the items assigned to the WEST and MARITIME domains could plausibly belong to the alternative domain, and the domain proportions reported here should be read as evidence-based estimates rather than exact partitions.

The separation between current-driven and directly dumped items relies on indirect condition indicators (weathering, biofouling, label integrity and buoyancy), which are semi-quantitative and respond to residence time, exposure and polymer type as well as to transport pathway. These indicators were evaluated during inspection at the level of country-of-origin batches rather than scored individually for every item, which limits the resolution at which within-batch heterogeneity can be resolved: items sharing a country of manufacture but differing in condition are treated alike under the present scheme, and a fully item-resolved condition survey would be required to refine the separation further. Lightly weathered items of distant origin were interpreted as maritime inputs, but recent release from a nearby coastal source, or rapid entrainment within a mesoscale eddy, can generate a comparable condition signal. The MARITIME contribution (approximately 39%), and hence MDS, should therefore be regarded as a well-supported estimate rather than a direct measurement; independent verification through vessel-density analyses based on Automatic Identification System data, Lagrangian drift simulations, or polymer-aging indicators would substantially strengthen this attribution.

Finally, the distance-weighting scheme underlying TCI is deliberately simple. It assigns three discrete classes (LOCAL = 1, BASIN = 2, GLOBAL = 3) rather than continuous transport distances or modeled travel times, and it treats the MARITIME domain as a global-scale input even though vessel-derived items may be released at any distance from the coast. TCI values are consequently sensitive to the weighting convention adopted and are best interpreted as a relative, internally consistent measure for comparison among sites analyzed under the same scheme, rather than as an absolute measure of transport distance. Testing the sensitivity of the index to alternative weighting schemes, and coupling the framework with particle-tracking simulations that resolve travel times explicitly, represent priority avenues for further development.

Implications for management and governance

The connectivity-driven nature of the assemblage provides a quantitative basis for translating scientific evidence into targeted management and governance actions. Connectivity-derived indices provide operational and quantifiable indicators that directly link system structure with policy-relevant intervention points. For instance, the high MDS (≈ 0.39) identifies maritime activities as a major input pathway, supporting the need for strengthened enforcement of international conventions such as MARPOL (International Convention for the Prevention of Pollution from Ships) and improved monitoring of shipping routes and waste reporting systems (Saliba et al., Reference Saliba, Frantzi and van Beukering2022; Galgani et al., Reference Galgani, Shim, Bessa, Piermarini, Rangel-Buitrago, Topouzelis and Gilardi2025). Similarly, the moderate SDI = 0.45 indicates that no single geographic domain controls the system, reinforcing the necessity of regional cooperation frameworks within interconnected marine basins (Kirschke et al., Reference Kirschke, van Emmerik, Nath, Schmidt and Wendt-Potthoff2023). The elevated TCI (≈ 2.41) further demonstrates that plastic pollution is governed by multi-scale processes, justifying coordinated international strategies addressing plastic production, transport and waste management (UNEP, 2021; Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2022).

At the corporate level, the low CDI ≈ 0.19 and high producer richness highlight a system of diffuse responsibility, where multiple producers contribute to environmental impacts. This structure limits the effectiveness of single-actor interventions and supports the expansion of EPR frameworks toward collective and transnational models (Cowger et al., Reference Cowger, Willis, Bullock, Conlon, Emmanuel, Erdle, Eriksen, Farrelly, Hardesty, Kerge, Li, Li, Liebman, Tangri, Thiel, Villarrubia-Gómez, Walker and Wang2024). In particular, the dominance of standardized, high-frequency items such as beverage bottles indicates that policy efforts should prioritize product-specific regulation, including reusable packaging systems, deposit-return schemes and harmonized design standards across markets (Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2019; Tian et al., Reference Tian, Zhong, Chen, Feng, Sun, Zhang, Shao, Liu and Hubacek2024).

The results of this study demonstrate clear limitations of local and site-based management approaches in addressing marine plastic pollution (Table 4). The fully transboundary nature of the assemblage, combined with its multi-source origin and connectivity-driven structure, indicates that a substantial portion of coastal litter originates outside the immediate management jurisdiction. Consequently, interventions focused solely on local waste management, while necessary, are insufficient to effectively reduce plastic accumulation in coastal environments influenced by regional and global transport processes (Jambeck et al., Reference Jambeck, Geyer, Wilcox, Siegler, Perryman, Andrady, Narayan and Law2015; Shumway and Ward, Reference Shumway and Ward2025).

Table 4.

Governance implications derived from connectivity-driven plastic systems

Table 4. Long description

The table consists of five columns: Scale, Key issue, Evidence from study, Implication, and Recommended action.

* Local scale: Key issue is external inputs dominate. Evidence includes T S I equals 1 and transboundary sources. Implication is local management is insufficient. Recommended action is to maintain local cleaning plus monitoring.

* Regional scale: Key issue is cross-border transport. Evidence includes W E S T dominance and Caribbean circulation. Implication is shared responsibility. Recommended action is regional agreements and data sharing.

* Global scale: Key issue is transoceanic inputs. Evidence is the T R A N S A T L A N T I C domain. Implication is that it is a global problem. Recommended action is an international treaty and global governance.

* Maritime scale: Key issue is direct sea-based inputs. Evidence is M D S is approximately 0.39. Implication is weak enforcement. Recommended action is to strengthen M A R P O L and monitoring.

* Corporate scale: Key issue is diffuse responsibility. Evidence includes low C D I and many producers. Implication is no single accountable actor. Recommended action is to strengthen E P R and global producer policies.

A regional cooperation framework is therefore essential, particularly within semi-enclosed basins such as the Caribbean. Given the strong role of regional circulation in redistributing plastics across national boundaries, coordinated monitoring, data sharing and joint mitigation strategies are required (Kirschke et al., Reference Kirschke, van Emmerik, Nath, Schmidt and Wendt-Potthoff2023). Regional agreements should aim to harmonize methodologies, standardize reporting systems and facilitate collective action among countries sharing connected marine systems, enabling more effective identification of priority sources and improving the efficiency of mitigation measures (Râpă et al., Reference Râpă, Cârstea, Șăulean, Popa, Matei, Predescu, Predescu, Donțu and Dincă2024).

At broader spatial scales, the observed transoceanic connectivity confirms that marine litter is inherently a global problem requiring coordinated international governance. The presence of inputs originating from distant regions, including transatlantic sources, demonstrates the need for regulatory frameworks that address plastic pollution across the full production–consumption–disposal chain (STAP, 2011; UNEP, 2021). Ongoing efforts toward a legally binding global agreement on plastic pollution represent a critical step, but their effectiveness will depend on incorporating scientific evidence on connectivity and transport dynamics (Williams and Rangel-Buitrago, Reference Williams and Rangel-Buitrago2022).

The significant contribution of the MARITIME domain further emphasizes the importance of strengthening maritime regulation and enforcement. Existing international conventions, such as MARPOL, provide a regulatory basis for controlling waste disposal at sea, yet enforcement remains inconsistent (Galgani et al., Reference Galgani, Shim, Bessa, Piermarini, Rangel-Buitrago, Topouzelis and Gilardi2025). Enhanced monitoring of shipping routes, improved reporting systems and stricter compliance mechanisms are necessary to reduce illegal or accidental dumping (Saliba et al., Reference Saliba, Frantzi and van Beukering2022). Targeted interventions within the shipping and fisheries sectors therefore represent a critical pathway for reducing marine litter inputs (Apete et al., Reference Apete, Martin and Iacovidou2024).

The diffuse corporate structure identified in this study underscores the need for strengthened producer responsibility frameworks. Given the globalized nature of production and consumption, such measures must be implemented through coordinated international approaches to effectively address the systemic drivers of plastic pollution (Bellou et al., Reference Bellou, Gambardella, Karantzalos, Monteiro, Canning-Clode, Kemna, Arrieta-Giron and Lemmen2021; UNEP, 2021; Rangel-Buitrago, Reference Rangel-Buitrago2025a).

Toward a quantitative framework for global studies

Marine litter has traditionally been interpreted within a local pollution framework, where accumulation patterns are primarily attributed to nearby sources and site-specific processes. While this perspective has been useful for baseline assessments, it is increasingly insufficient to explain the spatial distribution and persistence of plastics in the ocean. The results of this study support a broader conceptual shift toward understanding marine litter as a connectivity-driven system, where accumulation emerges from interactions operating across regional, basin and global scales.

Within this paradigm, plastic assemblages are not viewed as isolated deposits, but as components of a continuous and interconnected network, shaped by the combined influence of production systems, transport pathways and environmental filters. Ocean currents act as large-scale vectors that redistribute materials across vast distances (van Sebille et al., Reference van Sebille, Aliani, Law, Maximenko, Alsina, Bagaev, Bergmann, Chapron, Chubarenko, Cózar, Delandmeter, Egger, Fox-Kemper, Garaba, Goddijn-Murphy, Hardesty, Hoffman, Isobe, Jongedijk, Kaandorp, Khatmullina, Koelmans, Kukulka, Laufkötter, Lebreton, Lobelle, Maes, Martinez-Vicente, Morales Maqueda, Poulain-Zarcos, Rodríguez, Ryan, Shanks, Shim, Suaria, Thiel, van den Bremer and Wichmann2020), while maritime activities introduce direct inputs that further increase system complexity (UNEP, 2021). At the same time, the intrinsic properties of plastics (particularly their mobility and persistence) enable them to function as long-lived carriers of both material and associated contaminants across spatial scales (Andrady, Reference Andrady2011).

The concept of connectivity-driven anthropogenic assemblages provides a unifying framework to interpret these dynamics. In this view, assemblages represent the integrated outcome of multiple overlapping processes, including regional circulation, transoceanic transport and globalized consumption patterns. This perspective aligns with recent advances in marine litter research that emphasize long-range transport (Brabo et al., Reference Brabo, Martins, Andrades, Teixeira, do Nascimento, de Azevedo, Bezerra, Cavalcante, Cottens, Soares, de Oliveira Sousa, Mont’Alverne, Soares and Giarrizzo2024), source–sink decoupling (Ben-Haddad et al., Reference Ben-Haddad, Abelouah, Hajji, Rangel-Buitrago and Alla2023) and the role of ocean circulation in structuring debris distribution (Ramalhosa et al., Reference Ramalhosa, Monteiro, Rech, Gestoso, Álvarez, Gizzi, Parretti, Castro, Almeida, Jiménez, Ros, Cardoso, Lima, Caldeira, Robalo, Carlton and Canning-Clode2025). However, the present study extends these approaches by explicitly linking connectivity with hierarchical structure and quantitative metrics, demonstrating that plastic assemblages are both organized and measurable systems.

The results of this study provide the basis for the development of a standardized quantitative framework to investigate marine plastic pollution across regions. By integrating hierarchical classification, quantitative indices and connectivity-based source attribution, this approach moves beyond descriptive assessments toward a reproducible and comparable analytical system. Such a framework enables the systematic evaluation of plastic assemblages in terms of structure, diversity, source contribution and transport dynamics, facilitating cross-site and cross-regional comparisons.

One of the key strengths of this approach lies in its transferability. The combination of ATPL and ATPI does not depend on specific geographic conditions but can be applied to any coastal or marine environment where plastic litter is present. By maintaining a consistent classification system and standardized metrics, datasets from different regions can be directly compared, allowing the identification of global patterns in plastic distribution, corporate contribution and connectivity pathways. This represents a critical step toward building a global understanding of marine litter systems.

The incorporation of connectivity metrics further enhances the framework by explicitly accounting for transboundary processes, which are often overlooked in traditional studies. Indices such as source contribution, dominance and connectivity provide a quantitative basis to evaluate the relative importance of regional, maritime and long-distance inputs. This allows researchers to move from qualitative interpretations of transport to measurable connectivity structures, improving the robustness of source attribution and system analysis.

Importantly, this framework also supports the integration of scientific data with policy and management applications. By linking plastic assemblages to specific producers, product types and transport pathways, it becomes possible to identify priority intervention points across the production–consumption–disposal chain. In this context, ATPI metrics offer a valuable tool for evaluating corporate responsibility, monitoring trends over time and assessing the effectiveness of mitigation strategies.

The approach proposed in this study provides a foundation for the development of globally consistent monitoring and assessment systems, capable of capturing the complexity of marine plastic pollution. By combining classification, quantification and connectivity within a unified framework, it enables a shift toward integrated, data-driven analyses, supporting both scientific advancement and evidence-based decision-making at regional and global scales.