Plastic pollution is distributed patchily around the world’s oceans. Likewise, marine organisms that are vulnerable to plastic ingestion or entanglement have uneven distributions. Understanding where wildlife encounters plastic is crucial for targeting research and mitigation. Oceanic seabirds, particularly petrels, frequently ingest plastic, are highly threatened, and cover vast distances during foraging and migration. However, the spatial overlap between petrels and plastics is poorly understood. Here we combine marine plastic density estimates with individual movement data for 7137 birds of 77 petrel species to estimate relative exposure risk. We identify high exposure risk areas in the Mediterranean and Black seas, and the northeast Pacific, northwest Pacific, South Atlantic and southwest Indian oceans. Plastic exposure risk varies greatly among species and populations, and between breeding and non-breeding seasons. Exposure risk is disproportionately high for Threatened species. Outside the Mediterranean and Black seas, exposure risk is highest in the high seas and Exclusive Economic Zones (EEZs) of the USA, Japan, and the UK. Birds generally had higher plastic exposure risk outside the EEZ of the country where they breed. We identify conservation and research priorities, and highlight that international collaboration is key to addressing the impacts of marine plastic on wide-ranging species.
Plastic pollution harms marine life worldwide1, alongside other threats including fishing, climate change and invasive species2. Reports of entanglement and ingestion impacts are mounting1,3,4, but there are large gaps in our understanding, including about factors affecting plastic encounter, ingestion rates, mortality and population-level impacts4,5. Marine plastic is unevenly distributed6, accumulating in patches within ocean gyres and coastal regions7,8, and often drifting thousands of kilometres in ocean currents8,9. Likewise, marine life is patchily distributed10, and many species cross oceans and political boundaries11,12. With plastic production and waste generation continuing to increase13, identifying at-risk species and populations is crucial for targeting conservation action and research14,15,16 because the vulnerability of populations relates to exposure to a hazard, sensitivity to damage that impacts survival or reproduction, and the resilience of the population17.
Many seabird species are sensitive to plastic pollution; they frequently ingest plastic1, which can have lethal and sublethal impacts caused by chemical contamination18 and physical damage or blockages19. Numerous factors affect the amount of plastic accumulated by different species including foraging behaviour, at-sea distribution and gut morphology20,21,22. Among seabirds, albatrosses and petrels can contain particularly high loads of plastic ingested directly or within their prey1,20. Many species rarely regurgitate indigestible items, except when feeding their chicks23. Petrels are particularly sensitive because they retain plastic for long periods due to their gut morphology22, and small species (e.g., storm-petrels and gadfly petrels) can suffer greater physical damage or higher metabolic costs from ingesting plastic relative to larger species5. Petrels are a diverse group of 123 wide-ranging species that inhabit all the world’s oceans, making them good sentinels for ocean health2. Many populations are unlikely to be resilient to hazards because over half (64) are listed as globally Threatened or Near Threatened by the International Union for the Conservation of Nature (IUCN), including 16 Endangered and 12 Critically Endangered species2. Moreover, we know little about the status of many of their populations or if they are impacted by plastic2.
Assessing risk to petrel populations from plastic pollution requires a robust understanding of vulnerability to ingestion, for which exposure at sea is a key component14. Seabirds risk encountering plastic when they forage near sources associated with dense human populations24, fisheries25 and shipping lanes26, or in mid-ocean gyres where floating debris accumulates27,28,29. Exposure risk can be characterised by estimating contact between organisms and hazards, or their co-occurrence, and a key goal in ecological risk assessment is to consider variation in the amount of time spent by animals in different parts of their range30,31. Plastic exposure risk has not been previously quantified using methods that account for the time spent in areas of different densities of plastic pollution, but lightweight tracking devices have recently provided unprecedented detail about the movements of petrels of all sizes32, including the time spent in different foraging areas and across the annual cycle33.
Here, we estimate relative marine plastic exposure risk for 77 petrel species at a global scale by calculating the spatio-temporal overlap between modelled floating plastic density and the space-use of tracked birds14. To inform conservation action and future research, we compare exposure risk across populations, seasons (breeding and non-breeding), Exclusive Economic Zones (EEZ) and areas beyond national jurisdiction (the high seas), and found substantial variation. We identified areas of high risk of exposure to plastic debris in the Mediterranean and Black seas, the northeast Pacific, the northwest Pacific, the South Atlantic and the southwest Indian Ocean. Our results also reveal that Threatened species have greater exposure risk. Because marine debris and seabirds cross multiple political boundaries, our results emphasise that efforts to reduce the amount of plastic waste in the ocean should not only focus on areas of high exposure risk. Improved international cooperation and collaboration are needed to address this global threat.
Results and discussion
Plastic exposure risk for petrels
We analysed 1,736,880 tracked locations for 7137 adults of 77 petrel species (64% of species within Oceanitidae, Hydrobatidae and Procellariidae, excluding the two Macronectes species), from 148 populations in 27 countries and Antarctica, between 1995 and 2020 (mean = 2012). For each population, we calculated monthly 95% utilisation distributions (UDs) that estimate time spent by tracked petrels in 10 km grid cells (i.e., smoothed density of 12-hourly tracked locations; Fig. 1a), and combined monthly UDs into seasons (breeding or non-breeding). If data were available from multiple populations of a species, we created species UDs weighted by approximate population size. We calculated a geometric mean of global marine plastic densities estimated by three published models6,9,34 for micro- and macro-plastics (~0.333 mm–40 cm) combined for 2014 in 1 × 1° cells (Fig. 1b). We aggregated petrel UDs into 1 × 1° grid cells and created an all-species map by summing species UDs, weighting those tracked only in the breeding season and so not including the non-breeding part of the annual cycle at 0.5 (Fig. 1c). We divided the plastic and petrel grids by their respective cumulative sums so that the values of each global grid summed to one. We then multiplied each petrel UD by the plastic density to map spatial overlap as an indicator of estimated exposure risk14 (e.g., Figure 1d). Summing the values across cells provided an exposure risk score, which we multiplied by 106 to provide an easy-to-use scale; this gave us monthly population-level scores ranging from 0.0007 to 1091.
We ranked species by plastic exposure risk score (Fig. 2a), ranging from 0.003 to 549 (mean = 28.0; median = 4.9, interquartile range = 1.8–14.5). Of particular concern are the 19 species scoring over 15.3 (the score any species would receive if plastic was evenly distributed worldwide), indicating they mostly use areas with above-average plastic density. These species include the Critically Endangered Balearic shearwater Puffinus mauretanicus and Newell’s shearwater Puffinus newelli; the Endangered Hawaiian petrel Pterodroma sandwichensis; and the Vulnerable yelkouan shearwater Puffinus yelkouan, Cook’s petrel Pterodroma cookii and spectacled petrel Procellaria conspicillata (Fig. 2a). The proportion of total exposure risk within each IUCN Red List category differs from the proportion of tracked species within each category, with a greater percentage of the exposure risk shared among Threatened species, particularly Critically Endangered species (Fig. 2b). The 20 highest-scoring species had greatest plastic exposure risk in five areas, both in coastal regions (Mediterranean/Black Sea, northwest Pacific) and ocean gyres (northeast and northwest Pacific, South Atlantic, southwest Indian oceans; Figs. 1d, 2a). Plastic exposure risk was low in upwelling zones (Humboldt and Canary currents) and polar regions (Fig. 1d). For some species, scores differed greatly among populations (Fig. 2a). For example, European storm-petrels Hydrobates pelagicus breeding in the Mediterranean had much higher scores (306–534) than elsewhere (1.0–1.4; Supplementary Fig. 1). There was no long-term trend in exposure risk scores for populations tracked in the same months for more than three years (Supplementary Fig. 2). By using tracking data to estimate the relative density of regularised bird locations, instead of using only estimated presence or absence, we explicitly consider spatio-temporal variation in seabird distributions, thus providing more detail on global plastic exposure risk for a subset of species than an analysis based on range maps, which inferred different geographic hotspots of plastic exposure risk14.
Breeding and non-breeding season exposure risk
We calculated breeding and non-breeding plastic exposure risk scores for 107 populations of 60 species. The mean difference between seasons was 34.0, with little difference for most populations (median = 3.6), but substantial differences for some (maximum = 521.8; Fig. 3a). For example, Scopoli’s shearwaters Calonectris diomedea breed on Malta in the Mediterranean and migrate to the eastern Atlantic Ocean where they had a much lower plastic exposure risk score (30.0) than during the breeding season (496.2). In contrast, yelkouan shearwaters also breed on Malta (517.5), but had a higher score during non-breeding (937.7) when they disperse within the Mediterranean and migrate to the Black Sea (Fig. 3a–c). Seasonal contrasts also varied among populations of the same species. For example, scores for Cook’s petrels during non-breeding were much higher for birds breeding in northern New Zealand that migrate to the northeast Pacific (159.3), than those breeding in southern New Zealand that migrate to the Humboldt Current (0.8; Fig. 3a, d, e).
Exposure risk and ingestion
Plastic exposure risk, as indicated by our scores, is necessary but not sufficient for ingestion to occur and there are not yet enough suitable samples to quantify this process for most species. The amount of ingested plastic detected in seabirds is affected by foraging style, body size, tendency to regurgitate, gut morphology, prey type, age and breeding stage20,22,23,28. Few ingestion studies have used standardised protocols to sample different populations of the same species4. Furthermore, ingestion data are influenced by whether samples came from pellets26 or regurgitates18, or necropsies of birds that were found dead at a colony29 or on beaches35, recovered after attraction to light pollution36, bycaught in fisheries37, or taken for research28 or human consumption4. Nonetheless, studies that compared ingestion for different populations of the same species using the same methods control for these factors, and so can be compared to our exposure risk scores. For example, flesh-footed shearwaters Ardenna carneipes sampled in eastern parts of their breeding range contained significantly more plastic20, consistent with our higher scores during the non-breeding season for populations migrating to the northwest Pacific (New Zealand = 44.9; Lord Howe = 47.1) compared with those migrating to the eastern Indian Ocean (Western Australia = 13.6). Additionally, the Ecological Quality Objective for part of the North Sea target of <10% of northern fulmars Fulmarus glacialis containing ≥0.1 g of plastic was exceeded more in the North Sea than Arctic Canada38, mirroring our exposure risk scores for those tracked from the UK (1.4) and Canada (0.25). There are clear examples of high ingested plastic loads in high exposure risk areas in the Mediterranean37, northeast Pacific39 and southwest Indian Ocean36. However, plastic loads are both low and high in areas with low exposure risk40, indicating that birds may still be at risk while foraging in marine areas with low estimated plastic densities. Plastic has been ingested even by the species with the lowest exposure risk score of 0.003 (4% of 27 sampled snow petrels Pagodroma nivea, which forage around Antarctica, contained plastic40), indicating that the ubiquitous availability of plastic is concerning across all oceans worldwide, not only in areas where plastic aggregates.
Jurisdictions and policy
Plastic exposure risk for tracked petrels occurred mostly in the Mediterranean and Black Seas (Fig. 4a, b), where breeding European storm-petrels and Scopoli’s, yelkouan and Balearic shearwaters are at risk, with high plastic loads recorded37,41. Elsewhere, the high seas are used by 75 of our 77 tracked species, and accounted for 25% of global plastics exposure risk, mainly within oceanic gyres. The US EEZ accounted for a high proportion of the exposure risk, noticeably northeast of Hawai’i, followed by the EEZs of Japan, and the UK, mainly around the Overseas Territories of Tristan da Cunha and Bermuda (Fig. 4a, b). The New Zealand EEZ ranked highly despite low plastic levels due to the exceptionally high petrel occurrence and diversity. Moderate plastic exposure risk scores (0.15–1.00% of total) occurred in the EEZs of France, Australia, Brazil, Portugal, Mauritius, China, Russia, Argentina, Madagascar, Bahamas, and Mexico (Fig. 4a).
Our results indicate that mitigating plastic pollution in the breeding country’s EEZ alone would not adequately protect most species throughout the annual cycle. We identified links between the countries within which each tracked petrel population breeds (including overseas territories) and the jurisdictions where those populations were exposed to plastic (Fig. 4c). Exposure risk primarily occurred outside the breeding country’s EEZ (theoretical EEZ in the Mediterranean because actual EEZs are not clearly defined), except for 7 of the 29 highest-scoring populations (e.g., wedge-tailed shearwaters Ardenna pacifica in the USA, and streaked shearwaters Calonectris leucomelas in Japan). Of the 29 highest-scoring populations, 25 were exposed to plastic in multiple EEZs. For example, streaked shearwaters breeding in South Korea were exposed in China, Malaysia, the Philippines, South Korea, Indonesia and Vietnam (Fig. 4c). Exposure risk was greatest in the high seas for 15 of the 29 highest-scoring populations, particularly those breeding in the USA, New Zealand, UK, Brazil, Australia, France, and Mauritius (Fig. 4b). For each petrel population, we provide the percentage of exposure risk occurring in each EEZ and the high seas to facilitate targeting mitigation and policy efforts towards key areas (Supplementary Data 1).
Marine vertebrates and plastic debris are globally distributed and highly mobile, and cross political boundaries within and beyond national jurisdictions11. Therefore, mitigating plastic pollution from marine and terrestrial sources will require efforts targeted across multiple jurisdictions and the high seas42. International cooperation, collaboration, resource mobilisation and information exchange are key to addressing marine plastic pollution43 by limiting still-increasing plastic waste production13, improving waste management, and cleaning up existing plastic. The International Convention for the Prevention of Pollution from Ships (MARPOL) Annex V prohibiting plastic waste discharge from vessels entered into force 31st December 198844, but plastics from marine sources still affect seabirds26 and account for at least 22% of ocean plastics45. Ghost fishing gear is a priority because it presents deadly entanglement risk25 and food web contamination after degradation at sea. Pollution from vessels could be reduced with more resources and incentives for monitoring and managing waste, and enforcing MARPOL and local regulations, particularly among developing countries46. A coordinated approach for plastic waste management could be achieved, for instance, through a global-scale treaty on plastics43, which could operate in synergy with MARPOL and other relevant bodies and frameworks, such as the Convention on Biological Diversity, Convention on the Conservation of Migratory Species, Agreement on the Conservation of Albatrosses and Petrels, Regional Seas Conventions and Action Plans.
Greater use of standard methods for future ingestion studies would facilitate comparison and help identify the drivers of plastic ingestion4,47. The relationship between exposure risk, ingestion and impact could be examined by concurrently sampling ingested plastic and tracking movements41,48, and measuring physiological impacts. Interspecific differences could be clarified by systematically comparing plastic loads in species that have similar geographic ranges and exposure risk scores. Crucially, it is unclear for which species or populations plastic ingestion reduces survival or productivity and how much exposure they can tolerate; so, studies of population-level impacts and how to separate these from known causes of population declines will be vital2,5. Four species with high plastic exposure risk scores but no ingestion data in a recent review1 are key research priorities: Hawaiian petrel and streaked shearwater within the main high-exposure risk areas, and Bermuda petrel Pterodroma cahow and Desertas petrel Pterodroma deserta elsewhere. Comparable ingestion data from different tracked populations of the same species with contrasting migration patterns (e.g., Cook’s petrel; Fig. 3d, e) would be particularly valuable.
Our tracking data covered almost all of the world’s oceans and all ocean regions within the ranges of 70% of analysed species, broadly matching seabird biodiversity in general10, but also reflecting known spatial biases in research effort, notably towards the Atlantic Ocean and latitudes south of 40°S32 (see Supplementary Table 2 for spatial coverage gaps). Our study included tracking data for all four petrel species that breed in the Mediterranean, but we identified 14 species that occur in other high-exposure risk areas, making them priorities for tracking studies (Supplementary Table 3). Additionally, both petrel tracking and ingestion data are sparse in coastal waters around east and southeast Asia, where high plastic densities occur, and the South Pacific and North Atlantic gyres, where moderate plastic densities occur10,32 (Fig. 1). We identified priority species for future research in each of these regions (Supplementary Table 4). Sample sizes varied substantially among species, from 3 to 960 individuals (median = 35, mean = 93), so additional tracking for some species could be beneficial (Supplementary Data 2). Furthermore, tracking immature birds or adults when deferring breeding could reveal differences in exposure risk33. Our method could also be applied to global-scale, multi-species tracking datasets12 for other marine megafauna, such as turtles and marine mammals, for which plastic pollution is also a threat1.
Collecting more data on plastic density, identifying sources, and developing density models to provide better spatial coverage at a higher resolution would aid targeted mitigation strategies, and enable a better understanding of the effects of spatial scale on plastic exposure risk. The models that produced the plastic density estimates used in our analysis involved interpolating over wide areas, whereas observed plastic densities tend to be more patchy49. There were limited plastic data, particularly for southeast Asia6, where a recent survey recorded high plastic levels50. The South Pacific has a high petrel species richness, but few samples were used to inform the plastic density models6. The plastic density model estimates covered most of the Arctic and Antarctic oceans, but had more missing values near the poles than in other regions (Fig. 1b), although the Southern Ocean is not thought to contain much plastic6. However, plastic accumulates around Svalbard in the Arctic51, which although only important for northern fulmars among petrels, could affect other taxa. Marine species also feed at different depths and so it would be valuable to examine how plastic varies vertically52. Repeated plastic sampling across longer timescales would improve temporal matching between plastic and seabird data and allow investigations into long-term changes in plastic exposure risk53. We provide example versions of the code used to produce our results to facilitate future research on different tracking or plastics datasets54.
In brief, we collated tracking data for petrels and computed gridded utilisation distributions (UDs) at a monthly scale. We then combined gridded distributions of marine plastic density and multiplied them by the petrel UDs to map estimated exposure risk. For each map, we summed the plastic exposure risk values in all cells to provide a score representing relative estimated exposure risk. We combined maps and scores to investigate variation in exposure risk between breeding and non-breeding seasons, among populations and species, and across Exclusive Economic Zones (EEZs) and the high seas. Steps for processing and analysing the data are described in detail below and represented graphically in Supplementary Fig. 3. All data handling was carried out in R55 and R scripts are provided, along with example data and templates54.
Petrel tracking data collation and processing
We collated tracking data that were collected using Global Positioning System (GPS) loggers, Platform Terminal Transmitters (PTTs) and Global Location Sensor (GLS) loggers deployed on adult petrels (Table S1; Oceanitidae, Hydrobatidae and Procellariidae). We searched for published and unpublished tracking data for all petrel species between March and August 2020, excluding the two giant petrel species Macronectes giganteus and M. halli because our analyses focused on marine areas and they regularly feed on land56. We obtained data for 77 species (64% of the 121 target species) from the Seabird Tracking Database (www.seabirdtracking.org), ZoaTrack (www.zoatrack.org)57, Movebank (www.movebank.org)58, and individual researchers (represented by authors of this study or detailed in the Supplementary Acknowledgements). We collated 1,736,880 tracked locations for 7137 individuals tracked from 27 countries and Antarctica. Datasets varied in terms of number of colonies per species, and numbers of individuals, years, and months tracked per population (Supplementary Data 2) and species (Supplementary Data 3).
We standardised tracking datasets to contain the following fields in the same format: latitude, longitude, datetime, species, colony name, colony latitude, colony longitude and device type. For GLS, we removed locations around the equinoxes (March equinox: −21, +7 days; September equinox: −7, +21 days) as they are unreliable59, unless latitudes were estimated using additional information such as sea surface temperature prior to our analysis. For GPS and PTT data, we filtered locations for unrealistic speeds (>90 km/h), and visually checked maps and removed locations that were clear outliers. We removed locations within 5 km of the colony for GPS data or within 15 km of the colony for PTT data, but not for GLS locations due to large location error for these devices. We linearly interpolated and resampled GPS and PTT datasets to the sampling frequency for GLS of two locations per day.
We grouped data for each species into 148 breeding populations determined according to jurisdiction, the distance between colonies, and overlap in at-sea distributions based on the tracking data, i.e., if distributions overlapped substantially (at a 1 × 1° scale) and colonies are in close geographical proximity and in the same country, we considered colonies to belong to the same population.
Density of tracked petrel locations
For each population, we pooled all locations for all individuals across all years by month, and then removed months with fewer than five locations. For each month, we reprojected tracked locations onto a Lambert azimuthal equal area projection centred around the geometric mean of all locations. We estimated kernel densities of tracked locations to compute a 95% UD, a common home-range metric, which, because the sampling frequency was standardised, represented the estimated time spent by all tracked petrels in that population within that month. We used the adehabitatHR R package60, using a cell size of 10 km2 and a smoothing factor of 200 km (based on the magnitude of error in estimating locations from GLS33). We trimmed all cells that fell over land (Natural Earth land 1:10 m polygons version 5.1.1 downloaded from www.naturalearthdata.com/) because these species do not forage in terrestrial environments and it is extremely rare for them to travel over land, so any locations are most likely due to device error33. We then reprojected the resulting rasters back to a latitude and longitude projection (WGS84).
Of the 148 tracked populations, 108 (61 species) were tracked both in the breeding and non-breeding seasons. For these populations, we collated published information on the timing of breeding at a monthly scale (Supplementary Data 4) for each species or, where possible, each population. We also labelled months as breeding or non-breeding based on the tracking data. Locations were not always available for all months, with March and September often excluded from GLS datasets due to the uncertainty in light-based geolocation around equinoxes. We first calculated the distance between each location at sea and the breeding colony. For each population, we calculated the mean distance from the colony for each month, and a mean of those monthly means. If the mean for a month was greater than the population-specific mean across all months or if no individuals travelled within 200 km (chosen due to the approximate 200 km error common when using GLS devices) of the colony, this month was classified as non-breeding. To ensure there was only one breeding and one non-breeding season, if the classification of one month differed from the previous and following months, it was re-classified. We used published values except in cases when a month was labelled as breeding, but the tracking data showed that the subset of tracked birds did not attend the colony during that month, in which case, we used the label identified by the distance-to-colony method. Breeding and non-breeding months, therefore, do not necessarily represent the general phenology of the species, but instead reflect the behaviour (distance from the colony) of the majority of tracked individuals in that month. A sensitivity analysis showed that plastic exposure risk scores calculated using published breeding schedules were highly correlated with those estimated using the tracking data, Kendall’s tau = 0.98 (z = 13.879, p < 0.001) for the breeding season, and tau = 0.97 (z = 10.810, p < 0.001) for the non-breeding season.
Plastic density distribution
We used estimated global marine plastic density (count per km2) in 1 × 1° grid cells, from publicly available outputs from three published Lagrangian particle tracking models (Maximenko34, Lebreton9, and van Sebille6). The model estimates combined floating micro and macro-plastics from ~0.333 mm to 40 cm, with different size classes having similar estimated distributions7. Although petrels can ingest plastic flexible plastic pieces 40–60 cm long, they generally consume smaller pieces61. The three models estimated plastic density using records from ~12,000 surface trawls. They provided particularly good spatial coverage in the northeast Pacific, northwest Atlantic and Australian waters, but particularly poor coverage at the poles, the waters around Southeast Asia, the northwest Indian Ocean, and the South Pacific6. The models simulate the movement of plastic particles through multiple years and then create a static probability grid for a single time point (2014) based on where particles spent most time up until 2014 (equivalent to a utilisation distribution). We do not expect interannual variation in plastic distribution to be substantial in comparison to the spatial scale of between-season seabird movement because plastics travel passively, take decades to break down, and have been released throughout the study period. Each model uses the trawl data along with weather conditions, ocean circulation models, and plastic sources and sinks to inform the movement of plastic particles and predict the number of particles in each sampled and unsampled 1 × 1° grid cell. The Maximenko model assumes particles can wash ashore and originate from a uniform input across the ocean surface34, the van Sebille model assumes no sinks for plastic and plastics originate at the coast6, and the Lebreton model assumes no sinks for plastic and plastics are sourced from river mouths9. None of the models incorporate sinking through the water column52, ingestion by marine organisms1, or fragmentation processes. For each ocean basin and model, a prediction value was compared to observed plastic counts, providing regression coefficients used to scale the model plastics distribution and predict plastic concentrations within all cells6. Each model represents observed ocean plastic concentrations well6, with observations generally falling within 1–2 orders of magnitude around the model estimate. Further details on the methods used to model plastic density, including on how regression coefficients were used and validated, are provided in Maximenko et al.34, Lebreton et al.9, and van Sebille et al.6. Despite the variation in sampling effort, the model outputs generally agree with subsequent surveys in the Mediterranean62, southeast Pacific63 and southeast Asia50.
We took the geometric mean (as opposed to the arithmetic mean) of the Maximenko34, Lebreton9, and van Sebille6 models to avoid bias in our plastic density layer toward the highest estimate from any individual model because the models have log scale variability between their estimates. Additionally, because the ocean is in constant flux, concentrations at any given location are constantly changing53, assuming a lognormal distribution of concentrations through time, the geometric mean will be a better estimate of the central tendency and closer to the median concentration than the arithmetic mean64. The model outputs varied in spatial coverage in coastal and polar regions (Supplementary Fig. 4), and when one of the models did not have an estimate within a cell, we used the geometric mean of the other models, or the estimate from the only available model. If there was no estimate from any model, this was marked as NA, which occurred mostly in the Arctic and the Antarctic, and in some coastal areas where the marine area was less than the 1 × 1° grid size. The model outputs were centred around 180°E. Values in cells at 0–1°W were incorrectly estimated so these were imputed from the mean values in the three adjacent cells east and west (177–180°E and 1–4°W).
Plastic exposure risk scores
We aggregated the monthly 10 × 10 km petrel 95% UDs for each population33 onto the same 1 × 1° global grid of the plastic density data. All petrel UDs and the plastic density grid were divided by the respective cumulative sum for each grid so that the values of each entire raster grid summed to one. We estimated exposure risk as the mathematical product of the petrel and the plastic values in each grid cell14. This gives equal weight to the number of plastic pieces in each cell and the density estimate for bird tracking locations in each cell. We assume that estimated density of bird tracking locations at equal time intervals is strongly related to the time spent at risk of exposure to plastic debris, because areas where seabirds spend more time are very likely to be where foraging is concentrated65, as a result of area-restricted searching behaviour66,67,68. We then summed all cell values and multiplied all scores by 1,000,000 to reduce the number of decimal places to produce a single score for that month (ranging from 0.0007 to 1091). For comparison, we calculated a theoretical score of 15.3, which represents what the exposure risk score would be for any species if all global grid cells contained the mean plastic density (i.e., assuming that plastic was evenly distributed across the world’s oceans). We combined monthly grids to produce grids for each population, breeding or non-breeding season (if data were available for non-breeding months) and species. Scores for each population are the mean of all tracked months, and scores for each season are the mean of all months in that season (Supplementary Data 5). We used the mean to allow comparison between species with different numbers of tracked months. Maps for most populations are in Supplementary Fig. 5. For the 33 species for which we had multiple tracked populations, we searched for published population estimates (Supplementary Data 6). We calculated species-level scores as the mean of scores for each population weighted by the population size and multiplied by 0.5 if the population was only tracked during the breeding season (Supplementary Data 7).
We tested how robust our results were in relation to population size estimates, sampling frequency and tracking year. Population estimates for some species have large uncertainty, so we tested the correlation between species-level scores calculated with and without weighting by population size using Kendall’s tau because scores are not normally distributed. They were highly correlated (tau = 0.83; T = 483, p < 0.001), so our results are unlikely to be affected by uncertainty in population size estimates.
To investigate possible effects of sampling frequency, we reprocessed the tracking data without subsampling all datasets to 12-hourly intervals. We identified 44 populations for which all data were derived from GPS or PTT devices. For each track, we calculated the median interval between successive locations and recorded the maximum median for each population, and if this was less than 6 h, we regularised tracking locations at that frequency (intervals ranging from 1 min to 5 h, median = 1 h, mean = 82 min). We performed kernel density estimation with the higher-frequency datasets using a smaller 50 km smoothing factor33 for the remaining 39 populations and used them to calculate exposure risk scores for each population. The scores estimated using the higher and lower resolution data were highly correlated (tau = 0.90, T = 703, p < 0.001), so we conclude that 12-hour sampling intervals and 200 km smoothing parameter are sufficient for a study of this scale.
Birds were tracked between 1995 and 2020 with a mean tracking year of 2012. Among the 148 populations, 139 (94%) were tracked within 5 years of 2014 (2009–2019), the year for which plastic density was estimated. Given petrels are long-lived and generally faithful to breeding sites69 and foraging areas during both breeding and non-breeding seasons70,71,72,73, we assumed that distributions were unlikely to vary substantially across the study period. Data on long-term trends in plastic ingestion by seabirds have not shown substantial increases during the study period27,74,75. A subset of 13 populations had been tracked with geolocators for the same set of months across more than three years (Supplementary Fig. 2). For these, we calculated an exposure risk score for each year and then tested the effect of population and year using a generalised linear model with a Gamma distribution (due to positive continuous right-skewed response variable). We checked model fit by simulating residuals using the DHARMa R package76.
We recorded the most recent IUCN Red List assessment threat category77, where 36 species were Least Concern (LC), 9 Near Threatened (NT), 16 Vulnerable (VU), 10 Endangered (EN) and 6 Critically Endangered (CR). Red List status categories from the year each species was first tracked remained the same for 71 of the 77 species, and we used the most recent assessment for the 6 species for which changes have occurred. Three were genuine changes relating to altered threats or conservation action (Westland petrel Procellaria westlandica from VU in 2016 to EN in 2017; Chatham petrel from CR in 2008 to EN in 2009 to VU in 2015; yelkouan shearwater from LC in 2004 to NT in 2008 to VU in 2012), while three were not genuine changes because they related to improved evidence for assessment (flesh-footed shearwater from LC in 2012 to NT in 2016; streaked shearwater from LC in 2012 to NT in 2015; spectacled petrel from CR in 2005 to VU in 2007)77,78. We calculated the proportion of the total of all exposure risk scores attributed to species in each threat category.
Spatial patterns in plastic exposure risk
We used the ranked species scores to identify global-scale high -exposure risk areas by recording the region in which each species had the highest scores. We created an all-species map by summing results for each species, with those tracked in both breeding and non-breeding seasons given a weight of 1, while the 16 species that were tracked only in the breeding season were given a weight of 0.5 to avoid undue bias towards breeding colonies. We also divided the all-species distribution grid by the cumulative sum so that all values sum to one and multiplied this by the plastic density grid to produce an exposure risk map. We then overlapped this all-species map with EEZs and the high seas, obtained as an open-source polygon layer79. Because national jurisdictions in the Mediterranean are not yet clearly defined or are subject to dispute, we used theoretical EEZs, which are defined as 200 nautical miles from the coastline or the median point between two coastlines unless treaties and agreements have been submitted to the UN80. We calculated the proportion of the global risk of exposure to plastic for all petrels in each EEZ/theoretical EEZ and in the high seas. For joint regimes and overlapping claims, the score was divided evenly between the involved sovereigns. To record the links between the breeding country and the jurisdictions of plastic exposure risk, we calculated the proportion of plastic exposure risk for each population by country in each EEZ/theoretical EEZ and in the high seas11 (Supplementary Data 1).
Spatial coverage and research priorities
To assess spatial coverage and identify research priorities for tracked species, we compared the distribution of the tracking data for each species with the estimated range maps77. We assessed whether major populations (>1% of the global population or 200 pairs) of each tracked petrel species were missing from any of 10 major ocean areas (NW/NE/SW/SE Atlantic, NW/NE/SW/SE Pacific, Indian or Southern Oceans) according to the SeaVoX Salt and Fresh Water Body Gazetteer (https://www.marineregions.org/). Our tracking data covered all ocean regions within the published estimated ranges of 54 of the 77 species considered (70%). Our data compilation also revealed the main gaps in coverage for the remaining 23 species (Supplementary Table 2).
To identify research priorities for high exposure risk areas identified in this study, we used range maps to identify species or populations for which tracking data were not included in this study, but range maps indicated they may overlap (Supplementary Table 3). We recorded ingestion frequency of occurrence as the percentage of individuals found to contain plastic and the number examined as reported in Kühn & van Franeker1. We also carried out this process for areas for which plastic density is high and range maps showed that petrel species may use these areas, but no tracking data were available for our study (Supplementary Table 4).
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
The plastic exposure risk data generated in this study and the plastic density data used in this study are provided in the Supplementary Data files, available at https://github.com/BirdLifeInternational/petrels-plastics and have been deposited in the Zenodo database at https://doi.org/10.5281/zenodo.7852143. The seabird tracking data are available under restricted access because the data were collected for other purposes that vary between datasets and revealing the exact locations of sensitive species may put them at risk. Access can be obtained by making a request to the owners of each dataset using the mechanisms provided by each database. Zoatrack (https://zoatrack.org/) dataset IDs: 57, 93, 102–112, 159, 253, 254, 762, 817. Movebank (https://www.movebank.org/) dataset IDs: 944960474, 200628745, 241140274. SEATRACK (https://seapop.no/en/seatrack/) for relevant northern fulmar data. U.S. Geological Survey data release: https://doi.org/10.5066/P9NTEXM6. Seabird Tracking Database (https://www.seabirdtracking.org/) dataset IDs: 434, 438, 439, 448, 466, 467, 506–511, 517, 518, 554, 555, 561, 571, 607, 609, 610, 627, 628, 634, 635, 637, 639, 658, 659, 662, 663, 667, 668, 670, 672–678, 683, 684, 686, 694–696, 704–706, 708–715, 736, 741, 783–786, 788, 789, 826, 827, 829–831, 836–842, 844, 854, 858–872, 879, 883–886, 888–893, 900, 945, 946, 949, 951–954, 959–963, 966, 967, 970–983, 986–998, 1004, 1028, 1029, 1031–1033, 1055–1061, 1081, 1083, 1084, 1086–1091, 1120, 1121, 1140–1142, 1233–1236, 1238, 1239, 1258, 1259, 1279, 1280, 1282, 1285–1289, 1298, 1314, 1317, 1326, 1343–1347, 1360–1362, 1375, 1386, 1401, 1404, 1409, 1410, 1413–1415, 1422–1425, 1440, 1443, 1449, 1452, 1453, 1460, 1461, 1463, 1481, 1482, 1485–1488, 1494, 1497–1500, 1520–1523, 1541, 1544, 1546, 1549–1551, 1553–1558, 1562–1570, 1574–1577, 1579–1582, 1585–1592, 1594–1600, 1602, 1603, 1606–1608, 1610, 1618, 1619, 1621–1625, 1630, 1665, 1668–1672, 1690, 1711–1717, 1738, 1908–1923, 2036–2038, 2042, 2044–2046–2049, 2051–2056, 2059, 2060, 2063–2066. Source data are provided with this paper.
R code used to produce the analysis can be accessed at https://github.com/BirdLifeInternational/petrels-plastics with the version on the date of publication archived at https://zenodo.org/record/803386154.
Kühn, S. & van Franeker, J. A. Quantitative overview of marine debris ingested by marine megafauna. Mar. Pollut. Bull. 151, 110858 (2020).
Dias, M. P. et al. Threats to seabirds: a global assessment. Biol. Conserv. 237, 525–537 (2019).
Ryan, P. G., Moore, C. J., van Franeker, J. A. & Moloney, C. L. Monitoring the abundance of plastic debris in the marine environment. Philos. Trans. R. Soc. B Biol. Sci. 364, 1999–2012 (2009).
Provencher, J. F. et al. Quantifying ingested debris in marine megafauna: a review and recommendations for standardization. Anal. Methods 9, 1454–1469 (2017).
Senko, J. F. et al. Understanding individual and population-level effects of plastic pollution on marine megafauna. Endanger. Species Res. 43, 234–252 (2020).
van Sebille, E. et al. A global inventory of small floating plastic debris. Environ. Res. Lett. 10, 124006 (2015).
Eriksen, M. et al. Plastic pollution in the world’s oceans: more than 5 trillion plastic pieces weighing over 250,000 tons afloat at sea. PLoS One 9, e111913 (2014).
van Sebille, E., England, M. H. & Froyland, G. Origin, dynamics and evolution of ocean garbage patches from observed surface drifters. Environ. Res. Lett. 7, 44040 (2012).
Lebreton, L.-M., Greer, S. D. & Borrero, J. C. Numerical modelling of floating debris in the world’s oceans. Mar. Pollut. Bull. 64, 653–661 (2012).
Jenkins, C. N. & van Houtan, K. S. Global and regional priorities for marine biodiversity protection. Biol. Conserv. 204, 333–339 (2016).
Beal, M. et al. Global political responsibility for the conservation of albatrosses and large petrels. Sci. Adv. 7, eabd7225 (2021).
Block, B. A. et al. Tracking apex marine predator movements in a dynamic ocean. Nature 475, 86–90 (2011).
Borrelle, S. B. et al. Predicted growth in plastic waste exceeds efforts to mitigate plastic pollution. Science 369, 1515–1518 (2020).
Wilcox, C., van Sebille, E. & Hardesty, B. D. Threat of plastic pollution to seabirds is global, pervasive, and increasing. Proc. Natl. Acad. Sci. USA 112, 11899–11904 (2015).
Good, T. P., Samhouri, J. F., Feist, B. E., Wilcox, C. & Jahncke, J. Plastics in the Pacific: assessing risk from ocean debris for marine birds in the California current large marine ecosystem. Biol. Conserv. 250, 108743 (2020).
Provencher, J. F. et al. A horizon scan of research priorities to inform policies aimed at reducing the harm of plastic pollution to biota. Sci. Total Environ. 733, 139381 (2020).
Van Straalen, N. M. Biodiversity of ecological responses in animals. Netherlands J. Zool. 44, 1544–1556 (1994).
Lavers, J. L., Bond, A. L. & Hutton, I. Plastic ingestion by Flesh-footed Shearwaters (Puffinus carneipes): Implications for fledgling body condition and the accumulation of plastic-derived chemicals. Environ. Pollut. 187, 124–129 (2014).
Roman, L., Hardesty, B. D., Hindell, M. A. & Wilcox, C. A quantitative analysis linking seabird mortality and marine debris ingestion. Sci. Rep. 9, 3202 (2019).
Roman, L., Bell, E., Wilcox, C., Hardesty, B. D. & Hindell, M. Ecological drivers of marine debris ingestion in Procellariiform Seabirds. Sci. Rep. 9, 916 (2019).
Young, L. C., Vanderlip, C., Duffy, D. C., Afanasyev, V. & Shaffer, S. A. Bringing home the trash: do colony-based differences in foraging distribution lead to increased plastic ingestion in Laysan Albatrosses? PLoS One 4, e7623 (2009).
Ryan, P. G. How quickly do albatrosses and petrels digest plastic particles? Environ. Pollut. 207, 438–440 (2015).
Rodríguez, A., Rodríguez, B. & Carrasco, M. N. High prevalence of parental delivery of plastic debris in Cory’s Shearwaters (Calonectris diomedea). Mar. Pollut. Bull. 64, 2219–2223 (2012).
Lopes, C. S., de Faria, J. P., Paiva, V. H. & Ramos, J. A. Characterization of anthropogenic materials on yellow-legged gull (Larus michahellis) nests breeding in natural and urban sites along the coast of Portugal. Environ. Sci. Pollut. Res. 27, 36954–36969 (2020).
Bond, A. L. et al. Prevalence and composition of fishing gear debris in the nests of northern gannets (Morus bassanus) are related to fishing effort. Mar. Pollut. Bull. 64, 907–911 (2012).
Phillips, R. A. & Waluda, C. M. Albatrosses and petrels at South Georgia as sentinels of marine debris input from vessels in the southwest Atlantic Ocean. Environ. Int. 136, 105443 (2020).
van Franeker, J. A. & Law, K. L. Seabirds, gyres and global trends in plastic pollution. Environ. Pollut. 203, 89–96 (2015).
Spear, L. B., Ainley, D. G. & Ribic, C. A. Incidence of plastic in seabirds from the tropical Pacific, 1984-91: relation with distribution of species, sex, age, season, year and body weight. Mar. Environ. Res. 40, 123–146 (1995).
Lavers, J. L. & Bond, A. L. Ingested plastic as a route for trace metals in Laysan Albatross (Phoebastria immutabilis) and Bonin Petrel (Pterodroma hypoleuca) from Midway Atoll. Mar. Pollut. Bull. 110, 493–500 (2016).
Hope, B. K. An examination of ecological risk assessment and management practices. Environ. Int. 32, 983–995 (2006).
Chow, T. E., Gaines, K. F., Hodgson, M. E. & Wilson, M. D. Habitat and exposure modelling for ecological risk assessment: a case study for the raccoon on the Savannah River Site. Ecol. Modell. 189, 151–167 (2005).
Bernard, A., Rodrigues, A. S. L., Cazalis, V. & Grémillet, D. Toward a global strategy for seabird tracking. Conserv. Lett. 14, e12804 (2021).
Carneiro, A. P. B. et al. A framework for mapping the distribution of seabirds by integrating tracking, demography and phenology. J. Appl. Ecol. 57, 514–525 (2020).
Maximenko, N., Hafner, J. & Niiler, P. Pathways of marine debris derived from trajectories of Lagrangian drifters. Mar. Pollut. Bull. 65, 51–62 (2012).
Ryan, P. G. The incidence and characteristics of plastic particles ingested by seabirds. Mar. Environ. Res. 23, 175–206 (1987).
Cartraud, A. E., Le Corre, M., Turquet, J. & Tourmetz, J. Plastic ingestion in seabirds of the western Indian Ocean. Mar. Pollut. Bull. 140, 308–314 (2019).
Codina-García, M., Militão, T., Moreno, J. & González-Solís, J. Plastic debris in Mediterranean seabirds. Mar. Pollut. Bull. 77, 220–226 (2013).
van Franeker, J. A. et al. Monitoring plastic ingestion by the Northern Fulmar Fulmarus glacialis in the North Sea. Environ. Pollut. 159, 2609–2615 (2011).
Kain, E. C., Lavers, J. L., Berg, C. J., Raine, A. F. & Bond, A. L. Plastic ingestion by Newell’s (Puffinus newelli) and Wedge-tailed Shearwaters (Ardenna pacifica) in Hawaii. Environ. Sci. Pollut. Res. 23, 23951–23958 (2016).
van Franeker, J. A. & Bell, P. J. Plastic ingestion by petrels breeding in Antarctica. Mar. Pollut. Bull. 19, 672–674 (1988).
De Pascalis, F. et al. The hidden cost of following currents: microplastic ingestion in a planktivorous seabird. Mar. Pollut. Bull. 182, 114030 (2022).
Tiller, R. & Nyman, E. Ocean plastics and the BBNJ treaty—is plastic frightening enough to insert itself into the BBNJ treaty, or do we need to wait for a treaty of its own? J. Environ. Stud. Sci. 8, 411–415 (2018).
Borrelle, S. B. et al. Why we need an international agreement on marine plastic pollution. Proc. Natl. Acad. Sci. USA 114, 9994–9997 (2017).
Hagen, P. E. International community confronts plastics polluting from ships: MARPOL Annex V and the problem that won’t go away. Am. Univ. Int. Law Rev. 5, 425–496 (1990).
Morales-Caselles, C. et al. An inshore–offshore sorting system revealed from global classification of ocean litter. Nat. Sustain. 4, 484–493 (2021).
Karim, S. Implementation of the MARPOL convention in developing countries. Nord. J. Int. Law 79, 303–337 (2010).
Provencher, J. F. et al. Recommended best practices for plastic and litter ingestion studies in marine birds: collection, processing, and reporting. Facets 4, 111–130 (2019).
Hyrenbach, K. D. et al. Plastic ingestion by Black-footed Albatross Phoebastria nigripes from Kure Atoll, Hawai’i: linking chick diet remains and parental at-sea foraging distributions. Mar. Ornithol. 45, 225–236 (2017).
Goldstein, M. C., Titmus, A. J. & Ford, M. Scales of spatial heterogeneity of plastic marine debris in the northeast Pacific Ocean. PLoS One 8, e80020 (2013).
Tanhua, T., Gutekunst, S. B. & Biastoch, A. A near-synoptic survey of ocean microplastic concentration along an around-the-world sailing race. PLoS One 15, e0243203 (2020).
Cózar, A. et al. The Arctic Ocean as a dead end for floating plastics in the North Atlantic branch of the thermohaline circulation. Sci. Adv. 3, e1600582 (2017).
Egger, M., Sulu-Gambari, F. & Lebreton, L. First evidence of plastic fallout from the North Pacific Garbage Patch. Sci. Rep. 10, 7495 (2020).
Wilcox, C., Hardesty, B. D. & Law, K. L. Abundance of floating plastic particles is increasing in the western North Atlantic Ocean. Environ. Sci. Technol. 54, 790–796 (2020).
Clark, B. L., Pearmain, E. J., Carneiro, A. P. B., Cowger, W. & Dias, M. P. Code and data: global assessment of marine plastic exposure risk for oceanic birds. https://doi.org/10.5281/zenodo.7852143 (2023).
R Core Team. R: A Language and Environment for Statistical Computing. (R Foundation for Statistical Computing, Vienna, Austria, 2021).
Granroth-Wilding, H. M. V. & Phillips, R. A. Segregation in space and time explains the coexistence of two sympatric sub-Antarctic petrels. Ibis 161, 101–116 (2019).
Dwyer, R. G. et al. An open web-based system for the analysis and sharing of animal tracking data. Anim. Biotelemetry 3, 1–11 (2015).
Kranstauber, B. et al. The Movebank data model for animal tracking. Environ. Model. Softw. 26, 834–835 (2011).
Hill, R. D. & Braun, M. J. Geolocation by light level. In: electronic tagging and tracking in marine fisheries (eds. Sibert, J. R. & Nielsen, J. L.) 315–330 (Springer, 2001).
Calenge, C. The package ‘adehabitat’ for the R software: a tool for the analysis of space and habitat use by animals. Ecol. Modell. 197, 516–519 (2006).
Rodríguez, A., Ramírez, F., Carrasco, M. N. & Chiaradia, A. Seabird plastic ingestion differs among collection methods: examples from the short-tailed shearwater. Environ. Pollut. 243, 1750–1757 (2018).
Ruiz-Orejón, L. F., Sardá, R. & Ramis-Pujol, J. Floating plastic debris in the Central and Western Mediterranean Sea. Mar. Environ. Res. 120, 136–144 (2016).
Thiel, M. et al. Impacts of marine plastic pollution from continental coasts to subtropical gyres—fish, seabirds, and other vertebrates in the SE Pacific. Front. Mar. Sci. 5, 238 (2018).
Niku, S., Samaniego, F. J. & Schroeder, E. D. Discharge standard based on geometric mean. J. Water Pollut. Control Fed. 53, 471–473 (1981).
Warwick-Evans, V. et al. Time-in-area represents foraging activity in a wide-ranging pelagic forager. Mar. Ecol. Prog. Ser. 527, 233–246 (2015).
Fauchald, P. & Tveraa, T. Using first-passage time in the analysis of area restricted search and habitat selection. Ecology 84, 282–288 (2003).
Paiva, V. H., Geraldes, P., Ramírez, I., Garthe, S. & Ramos, J. A. How area restricted search of a pelagic seabird changes while performing a dual foraging strategy. Oikos 119, 1423–1434 (2010).
Clay, T. A., Oppel, S., Lavers, J. L., Phillips, R. A. & Brooke, M. D. L. Divergent foraging strategies during incubation of an unusually wide-ranging seabird, the Murphy’s petrel. Mar. Biol. 166, 8 (2019).
Bried, J. & Jouventin, P. Site and mate choice in seabirds: an evolutionary approach. In: Biology of Marine Birds (eds. Schreiber, E. A. & Burger, J.) 263–305 (CRC Press, 2002).
Ramírez, I. et al. Conservation implications of consistent foraging and trophic ecology in a rare petrel species. Anim. Conserv. 19, 139–152 (2016).
Delord, K. et al. Individual consistency in the non-breeding behavior of a long-distance migrant seabird, the grey petrel Procellaria cinerea. Mar. Ornithol. 47, 93–103 (2019).
Navarro, J. & González-Solís, J. Environmental determinants of foraging strategies in Cory’s shearwaters Calonectris diomedea. Mar. Ecol. Prog. Ser. 378, 259–267 (2009).
Raymond, B. et al. Shearwater foraging in the Southern Ocean: the roles of prey availability and winds. PLoS One 5, e10960 (2010).
Ryan, P. G. Seabirds indicate changes in the composition of plastic litter in the Atlantic and south-western Indian Oceans. Mar. Pollut. Bull. 56, 1406–1409 (2008).
Ryan, P. G., Pichegru, L., Perold, V. & Moloney, C. Monitoring marine plastics—will we know if we’re making a difference? S. Afr. J. Sci. 116, #7678 (2020).
Hartig F. DHARMa: residual diagnostics for hierarchical (multi-level/mixed) regression models. R package version 0.2.7. https://CRAN.R-project.org/package=DHARMa (2020).
BirdLife International. IUCN red list for birds [WWW Document]. URL. http://www.birdlife.org, Accessed date: 7 June 2021 (2019).
Butchart, S. H. M. et al. Measuring global trends in the status of biodiversity: red list indices for birds. PLoS Biol. 2, e383 (2004).
Flanders Marine Institute. Maritime boundaries geodatabase: maritime boundaries and exclusive economic zones (200NM), version 11. Available online at https://www.marineregions.org/10.14284/386. (2019).
MedPAN & SPA/RAC. The 2016 status of Marine Protected Areas in the Mediterranean. (eds. Meola B. & Webster C.) SPA/RAC & MedPAN. Tunis (2019).
B.L.C., C.H., and A.M. were funded by the Cambridge Conservation Initiative’s Collaborative Fund sponsored by the Prince Albert II of Monaco Foundation. E.J.P. was supported by the Natural Environment Research Council C-CLEAR doctoral training programme (Grant no. NE/S007164/1). We are grateful to all those who assisted with the collection and curation of tracking data. Further details are provided in the Supplementary Acknowledgements. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.
The authors declare no competing interests.
Peer review information
Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Clark, B.L., Carneiro, A.P.B., Pearmain, E.J. et al. Global assessment of marine plastic exposure risk for oceanic birds. Nat Commun 14, 3665 (2023). https://doi.org/10.1038/s41467-023-38900-z
This article is cited by
Legacy oceanic plastic pollution must be addressed to mitigate possible long-term ecological impacts
Microplastics and Nanoplastics (2023)