Aim: While marine debris is entering the ocean at unprecedented rates, little is known about how biological communities engage with this introduced substrate. This study analyzes the community composition of debris-associated fishes and explores the role of plastics in transferring fish across biogeographic barriers.

Location: Pacific Ocean, including North Pacific Garbage Patch.

Results: Fish communities around marine debris have low diversity, individual debris items have highly dissimilar species compositions, and species composition is affected by horizontal debris item size. Debris items in Group I have significantly higher fish species diversity and differ in community composition from Group II. We observed four species (Abudefduf vaigiensis, Histrio histrio, Oplegnathus punctatus, and Petroscirtes spp.) surviving east of the Hawaiian Islands, where they have not been previously reported. While we have not yet detected juveniles of these “break through” species east of the Islands, the earlier transport by debris of A. vaigiensis and H. histrio to Hawaii in the 1990s and their subsequent successful establishment there suggest that these Indo-Pacific fish have found hospitable thermal regimes east of the Hawaiian Islands matching their original reproductive ranges.

Main conclusion: We suggest that marine debris acts as a mechanism for fish to cross biogeographic barriers, such as the Eastern Pacific Barrier. While natural flotsam has long been a potential vector, long-lived and increasingly abundant plastic debris may set the stage for altering fish biogeography.

Five cruises were conducted between 2018 and 2020 by The Longest Swim (TLS), The Vortex Swim (TVS), Greenpeace (GP), and Ocean Voyages Institute (OVI, 2019 and 2020) (Table S1, Fig. 1). Each cruise had different research goals and photographic sampling methodologies; cruise tracks thus do not necessarily represent equal sampling intensity. In chronological order, TLS was a 189-day voyage from Japan to Hawaii; GP a 104-day cruise from Mexico to Hawaii; OVI-2019 a 31-day out-and-back cruise between Hawaii and the NPGP; TVS a 71-day cruise from Hawaii to San Francisco, and OVI-2020 a 59-day out-and-back cruise between Hawaii and the NPGP. Thus, TLS included samples from both the Western Pacific Ocean (west of 180 degrees longitude) and the Eastern Pacific Ocean; all other cruises were restricted to the Eastern Pacific Ocean.

We analyzed approximately 1,500 underwater photographs of marine debris taken during the cruises. In addition, we analyzed two videos, which were processed into individual images using Adobe Photoshop at 3 frames per second. Together, 49 debris items were analyzed. Spatial and temporal data (precise location and dates of photographs) were available for TVS and OVI cruises; precise data were not available for the other cruises (Table S1).

​​We recorded debris item attributes as follows:

Type: hollow (barrels, bins, boats, bottles, boxes, and pallets; total number of hollow items, n = 10), solid (buoys, spools/reels, and plastic pieces; n = 12), and nets (n = 21)

Size: using objects of estimated known size (such as a diver) for comparison, we recorded horizontal (item width) and vertical (penetration below the water’s surface) sizes as: small (20-50 cm; horizontal n=16, vertical n=15), medium (50-500 cm; horizontal n=15, vertical n=24), large (500-1000 cm; horizontal n=8, vertical n=3), and mega (1000+ cm; horizontal n=4, vertical n=1)

Material: hard plastic (n = 17), soft plastic (styrene; n = 3), wood (n = 1), plastic fabric (tarps; n = 5), rope (n = 14), and plastic and rope (mixed material items; n = 3).

Color: dark (brown; n = 4), light (white, yellow, light blue, gray, tan, or clear; n = 25), and multicolor (multiple colors, including dark and light components; n = 14)

Biofouling:  the presence (n = 26) or absence (n = 17) of the pelagic gooseneck barnacle Lepas, an ocean rafter and often foundation species for flotsam-associated invertebrate communities.

Only items with associated fish (n = 43) were analyzed. The six excluded items were from two cruises (GP and TVS) and included two bottles and four nets ranging from small to large, consisting of plastic and rope; five of these items had no Lepas biofouling.

Live fish were identified to the lowest taxonomic level possible. Dead fish (n = 4) were not recorded. Most individual debris items had multiple photos. Fish species diversity and maximum abundance per species were quantified for each debris item. To determine maximum abundance and reduce the likelihood of double counting, we recorded the maximum number of each fish species seen in a photo within a given item’s photoset. Fish identifications were verified or made by ichthyologists Giacomo Bernardi and Bruce Mundy.