Marine Biology

, Volume 161, Issue 6, pp 1441–1453

Relationship of diversity and habitat area in North Pacific plastic-associated rafting communities

Authors

    • Scripps Institution of OceanographyUniversity of California San Diego
    • California Sea Grant
  • Henry S. Carson
    • University of Hawaii at Hilo
    • Washington Department of Fish and Wildlife
  • Marcus Eriksen
    • 5 Gyres Institute
Original Paper

DOI: 10.1007/s00227-014-2432-8

Cite this article as:
Goldstein, M.C., Carson, H.S. & Eriksen, M. Mar Biol (2014) 161: 1441. doi:10.1007/s00227-014-2432-8

Abstract

Plastic and other anthropogenic debris (e.g., rubber, tar) augment natural floating substrates (e.g., algal rafts, pumice) in the open ocean, allowing “islands” of substrate-associated organisms to persist in an otherwise unsuitable habitat. We examined a total of 242 debris objects collected in the eastern Pacific in 2009 and 2011 (32–39°N, 130–142°W) and the western Pacific in 2012 (19–41°N, 143–156°E). Here, we ask: (a) What taxa are associated with plastic rafts in the North Pacific? and (b) Does the number of taxa associated with plastic debris vary with the size of the debris “island?” We documented 95 rafting taxa from 11 phyla. We identified several potentially invasive plastic-associated rafting taxa, including the coral pathogen Halofolliculina spp. In concordance with classic species–area curves, the number of rafting taxa was positively correlated with the size of the raft. Our findings suggest that diversity patterns on plastic debris are compatible with the concept of island biogeography.

Introduction

Naturally occurring floating objects in the pelagic environment have long played host to a suite of specialized species (Thiel and Gutow 2005a). These substrates, such as floating algae, pumice, and wood, provide transport and habitat for benthic organisms (Donlan and Nelson 2003; Thiel and Gutow 2005b; Bryan et al. 2012). In recent decades, natural rafts have been augmented by anthropogenic debris comprised primarily of non-biodegradable plastic polymers such as hard thermoplastic, foam, synthetic rubber, and fiberglass (Derraik 2002; Barnes et al. 2009).

Plastic debris was first detected in the open ocean in the early 1970s (Carpenter and Smith 1972; Venrick et al. 1973; Wong et al. 1974) and has now been observed all over the world (Thiel and Gutow 2005a). Plastic enters the marine environment through improper disposal (e.g., litter) or accidental loss (e.g., fishing gear; U.S. Environmental Protection Agency 2011). Debris from land-based sources is most common near highly populated areas, while debris from marine sources is most common on remote shores (Hammer et al. 2012). However, as debris is exposed to UV light and physical weathering, it fragments into small pieces, termed microplastics, that are frequently less than 5 mm in diameter (Andrady 2011; Hidalgo-Ruz et al. 2012). Microplastics now comprise the vast numerical majority of debris in the ocean (Goldstein et al. 2013) though larger objects that can support a more extensive rafting community are far from uncommon (Titmus and Hyrenbach 2011, Ryan 2013).

The composition of the rafting assemblage depends on the type of object, its stability, and the supply of propagules (Thiel and Gutow 2005b). In general, artificial substrates do not host the same communities as natural substrates (Tyrrell and Byers 2007; Pister 2009; but see Bravo et al. 2011). In the case of floating objects, biotic rafts (e.g., wood, detached kelp) do not float for as long as abiotic rafts (e.g., plastic, tar, pumice), but do provide a food source for rafting organisms, and therefore may be more successful at transporting a variety of species (Donlan and Nelson 2003; Thiel and Gutow 2005b). Items with a complex surface (e.g., pumice, macroalgae holdfasts) may provide better habitat than items with a smooth surface (e.g., plastic bottles; Thiel and Gutow 2005b). The rotational stability of the rafting object may also affect the diversity of the attached assemblage—pieces with fewer changes in orientation have greater species richness and cover than less stable pieces (Bravo et al. 2011; Bryan et al. 2012). Fouling also increases the specific density of the raft, which may cause sinking in the water column and potentially a subsequent rise to the surface if fouling organisms die or are removed by predators (Ye and Andrady 1991; Moret-Ferguson et al. 2010), though fouling can also help maintain the positive buoyancy of porous rafts (e.g., pumice, foam) by reducing gas permeability (Bryan et al. 2012). Lastly, the physical environment around the raft, such as distance from shore or water temperature, may be more significant to rafting species composition than characteristics of the raft itself (Clarkin et al. 2012). For example, rafts that were colonized in coastal waters may have different species composition than rafts colonized at sea (Astudillo et al. 2009), and the diversity of the rafting community may be enhanced by encounters with larval sources such as islands, reefs, and other shallow-water habitats (Bryan et al. 2012).

Diversity patterns in open-ocean rafting assemblages, particularly plastic-associated assemblages, are largely unexplored. In many ecosystems, diversity is predicted by the species–area curve, in which the number of species increases as a function of available surface area, though the shape of this curve has been a matter of some debate (He and Legendre 1996). Species–area curves can be explained by the concept of island biogeography, which predicts that species diversity is a balance between arrival of species through migration and the loss through extinction, both processes that vary with available habitat area (MacArthur and Wilson 1963). While species–area relationships are one of the most widely observed patterns in ecology, there are exceptions, such as the “small-island effect,” in which the areas of the ecosystems observed are all too small for a diversity pattern to be detected (Lomolino 2000). It is unknown whether plastic-associated rafting assemblages follow the species–area pattern, particularly since organisms with certain life history traits, such as suspension feeding, are more likely to be successful rafters, especially on abiotic substrata such as plastic (Thiel and Haye 2006).

The light weight and durability of plastic make it a vector for the transport of non-indigenous species. For example, a piece of flotsam with traces of tropical biota, including self-fertilizing corals, was recently discovered in the Netherlands (Hoeksema et al. 2012), and Southern Ocean bryozoans in reproductive condition were found on a beached packing band in Antarctica (Barnes and Fraser 2003). Benthic organisms such as bryozoans, barnacles, and hydroids are commonly found on plastic debris (Aliani and Molcard 2003; Barnes and Milner 2005; Farrapeira 2011). The particular vulnerability of island ecosystems to invasions and the ubiquity of plastic debris on the mid-Pacific islands makes lateral transport of fouling species a matter of particular concern in the North Pacific (McDermid and McMullen 2004). Recently, a non-native hydroid and two ascidians were recorded from debris collected in the Northwest Hawaiian Islands (Godwin et al. 2008). However, most studies have examined beached material, not in situ debris (Winston et al. 1997; Barnes 2002; Barnes and Fraser 2003; Barnes and Milner 2005; Hoeksema et al. 2012; but see Astudillo et al. 2009).

Even when debris does not carry organisms to distant shorelines, debris can provide abundant habitat to fouling organisms. In coastal areas, abundance of floating macroalgae varies between 1 and 1,000 items km−2, occasionally even exceeding values of 10,000 items km−2 (Thiel and Gutow 2005a). High densities of abiotic substrates can also occur in the open ocean—one eruption in Tonga was estimated to release over 2.5 × 1012 individual pumice clasts, more than 50 % of which were inhabited by rafting organisms (Bryan et al. 2012). Floating microplastic debris in the subtropical gyres can also reach high densities, such as a median of 425,000 items km−3 in the North Pacific subtropical gyre (Goldstein et al. 2012) and a mean of 20,328 items km−2 in the North Atlantic subtropical gyre (Law et al. 2010). This increase in habitat has the potential to expand populations of open-ocean rafting species, such as gooseneck barnacles (Whitehead et al. 2011) and oceanic insects (Goldstein et al. 2012; Majer et al. 2012).

In this study, we asked: (a) What taxa are associated with drifting plastic in the North Pacific? (b) Does the number of taxa associated with plastic debris vary with the size of the debris “island?”

Methods

Samples were collected on three cruises, the 2009 Scripps Environmental Accumulation of Plastics Expedition (SEAPLEX), the 2011 Algalita Eastern North Pacific Gyre Expedition, and 2012 Western North Pacific Gyre Expedition (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-014-2432-8/MediaObjects/227_2014_2432_Fig1_HTML.gif
Fig. 1

Map of sampling locations and photographs of representative plastic debris. a Locations of debris collection in 2009 (circles), 2011 (triangles), and 2012 (diamonds). b Small plastic fragments intermixed with the chondrophore Velella velella; collected August 10, 2009, 03:57 GMT, 32° 37.91′N, 140° 18.61′W. c Medium plastic fragments with 15-cm ruler; collected August 11, 2009, 07:30 GMT, 32° 54.99′N, 140° 19.81′W. d A large tangle of intermixed fishing-related nets, ropes, and buoys along with other hard plastic debris (“rope clump”); collected May 12, 2012, 03:00 GMT, 22° 13.35′N, 155° 21.17′E

For the 2009 samples, floating debris items were opportunistically collected by dip net (39 cm length × 33.5 cm width, mesh 1 mm). If possible, the entire piece of debris with attached fauna was preserved in either 5 % formalin buffered with sodium borate or 95 % ethanol. An attempt was made to preserve portions of most samples in both preservatives to allow for both morphological and molecular studies. When the item was too large to be preserved, the item was either subsampled (e.g., portions of a tarp were cut and preserved) or the fauna were removed and preserved separately (e.g., in the case of a large fishing buoy). A subset of plastics collected using a standard manta net (0.86 × 0.2 m) with 333-μm mesh (Brown and Cheng 1981), towed for 15 min at 0.7–1 m s−1, were also included in this study. For smaller Manta-net-collected debris particles, 50 % aliquots of the net-collected samples were analyzed. Since splitting samples cause the less abundant larger debris items to be undersampled, all Manta-net-collected objects with a diameter of greater than 2 cm were included in this study. For this reason, we found it practical to use the 2-cm cutoff to divide “fragments” from larger objects.

On the 2011 and 2012 expeditions, debris items were collected by dipnet during timed debris observation periods or opportunistically during other daylight sightings. Debris items were inspected for attached organisms immediately and then taken to shipboard laboratory for microscope inspection, photography, and preservation. If possible, the entire piece of debris with attached fauna was preserved in 5 % formalin as on the 2009 cruise. Where not possible, organisms were removed and preserved separately. Three floating masses of nets, rope, and entangled debris were sampled differently. In 2011, the net mass was examined by divers in the water, who noted associated fishes, collected fouling organisms opportunistically, and subsampled the materials for inspection on board. In 2012, both net masses were inspected in the water first and then hauled on board and dissected on deck for a more thorough collection of fouling organisms.

In the laboratory, objects from the 2009 expedition were examined for rafting fauna under a Wild M-5 dissecting microscope. The preservative was also filtered through 150-μm Nitex mesh to retain non-attached biota. During the 2011 and 2012 expeditions, organisms and small debris were inspected and photographed on board using a Dino-Lite Premiere digital microscope. All objects were photographed with in situ size references. Two-dimensional surface area was digitally measured using the NIH ImageJ software (Rasband 2012) and calibrated against manual measurements. Because of the flattened shape of most debris objects, we approximated total surface area by multiplying two-dimensional surface area by a factor of two. It should be noted that this approach substantially underestimates the total surface area of complex structures such as rope clumps and net balls.

Later identification of preserved specimens in the laboratory was made using dissection or compound microscopes. All fauna were identified to the lowest possible taxonomic level. When objects with different taxonomic resolutions were compared, taxa were collated to comparable levels. For example, Lepas pacifica, L. anatifera, and Lepas spp. were counted as one taxon. To determine whether taxa had previously been documented as rafting, we first consulted the comprehensive lists of rafting taxa given in Thiel and Gutow (2005b). If a given taxa were not listed in Thiel and Gutow (2005b), we conducted a literature search to determine whether we could find other documentation of rafting in the taxa. If we could not find such documentation, the taxa were listed as “not previously documented as rafting.” A complete list of debris locations and associated taxa is given in Supplemental Table 1. We determined the feeding type of each taxon from our own biological knowledge and from reference to the literature as necessary. When the feeding ecology of a specific taxon was unknown, we assigned it to the most probable feeding type. For example, an unidentified hydroid was classified as a “suspension feeder.”

Statistics and figures were generated with the R statistical environment, version R-2.13.1 (R Development Core Team 2012). Debris object areas were log-transformed for ease of display. We used the chi-square test to test whether rafting assemblage traits (e.g., phyla composition, feeding type) varied between cruises and years. We used Kendall’s tau rank correlation coefficient to measure the strength of dependence between debris size and number of taxa. For one analysis, both taxa and debris area were linearized using log transformations so that the relationship could be more easily visualized, and analyzed using linear regression (Lomolino 2000) (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-014-2432-8/MediaObjects/227_2014_2432_Fig2_HTML.jpg
Fig. 2

Photographs of assorted debris. aLepas barnacles growing on buoy; collected June 22, 2012, 02:00 GMT, 29°11.9′N, 170°35.2′E. bLepas barnacles growing on a rope; collected May 14, 2012, 20:00 GMT, 26°26.56′N, 152°07.44′E. c Two Lepas barnacles and membraniporid bryozoans growing on a toothbrush handle, collected June 17, 2012, 02:00 GMT, 15°26.3′N, 150°30.0′E. d A juvenile triggerfish C. maculata found associated with a bleach bottle, collected May 12, 2012, 21:00 GMT at 22°11.06′N, 155°22.07′E. e Close view of fragment showing folliculinid ciliates and other organisms; collected May 10, 2012, 22:15 GMT, 19°53.05′N, 155°04.22′E. f Close view of folliculinid ciliates showing the two pericytostomial wings extending from the lorica, collected May 10, 2012, 22:15 GMT, 19°53.05′N, 155°04.22′E

Results

We examined a total of 242 debris objects and identified 95 associated rafting taxa (Table 1). The debris comprised 66 % rigid plastic fragments less than 2 cm in diameter, 21 % rigid plastic fragments or objects ranging from 2 to 100 cm in diameter, 7 % rope clumps, 3 % flexible substrates (e.g., tarps), and 3 % expanded foam (e.g., “Styrofoam”). Debris substrate area ranged from 2.54 × 10−7 to 15 m2, with a median of 1.18 × 10−4 m2.
Table 1

Rafting taxa

Phylum

Class

Order

Finest taxon identified

Year observed

Previously documented as rafting

Annelida

Polychaeta

Aciculata

Eunice spp.

c

1

Annelida

Polychaeta

Amphinomida

Amphinome rostrata

c

1

Annelida

Polychaeta

Amphinomida

Hipponoe gaudichaudi

a, b

1

Annelida

Polychaeta

Phyllodcida

Halosydna spp.

b

N

Annelida

Polychaeta

Phyllodocida

Nereididae

c

1

Annelida

Polychaeta

Phyllodocida

Nereis spp.

c

1

Annelida

Polychaeta

Phyllodocida

Phyllodocidae

c

1

Annelida

Polychaeta

Sabellida

Salmacina spp.

c

N

Annelida

Polychaeta

Sabellida

Subfamily Serpulinae

c

1

Annelida

Polychaeta

Sabellida

Subfamily Spirorbinae

a, c

1

Arthropoda

Malacostraca

Amphipoda

Caprella spp.

a, c

1

Arthropoda

Malacostraca

Amphipoda

Elasmopus spp.

a

1

Arthropoda

Malacostraca

Amphipoda

Hyalidae

a

1

Arthropoda

Malacostraca

Amphipoda

Isaeidae

b

N

Arthropoda

Malacostraca

Amphipoda

Pleustidae

c

N

Arthropoda

Malacostraca

Amphipoda

Stenothoidae

a

1

Arthropoda

Malacostraca

Amphipoda

Suborder Gammaridea

c

1

Arthropoda

Malacostraca

Decapoda

Chorilia spp.

c

N

Arthropoda

Malacostraca

Decapoda

Superfamily Majoidea

c

1

Arthropoda

Malacostraca

Decapoda

Megalopae

b

1

Arthropoda

Malacostraca

Decapoda

Palaemon affinis

c

1

Arthropoda

Malacostraca

Decapoda

Pilumnus spp.

c

N

Arthropoda

Malacostraca

Decapoda

Plagusia spp.

c

1

Arthropoda

Malacostraca

Decapoda

Plagusia squamosa

a

1

Arthropoda

Malacostraca

Decapoda

Planes cyaneus

a, c

1

Arthropoda

Malacostraca

Decapoda

Planes minutus

a

1

Arthropoda

Malacostraca

Decapoda

Planes spp.

b, c

1

Arthropoda

Malacostraca

Isopoda

Cirolanidae

a

1

Arthropoda

Malacostraca

Isopoda

Idotea spp.

a, b, c

1

Arthropoda

Malacostraca

Isopoda

Sphaeromatidae

a

1

Arthropoda

Maxillopoda

Harpacticoida

Harpacticoida

a

1

Arthropoda

Maxillopoda

Kentrogonida

Heterosaccus spp.

c

N

Arthropoda

Maxillopoda

Lepadiformes

Barnacle cyprids

a

1

Arthropoda

Maxillopoda

Lepadiformes

Lepas anitifera

a, c

1

Arthropoda

Maxillopoda

Lepadiformes

Lepas pacifica

a

1

Arthropoda

Maxillopoda

Lepadiformes

Lepas spp.

a, b, c

1

Arthropoda

Maxillopoda

Sessilia

(Amphi)balanus amphitrite

b

1

Arthropoda

Maxillopoda

Sessilia

Chthamalus spp.

c

N

Arthropoda

Maxillopoda

Sessilia

Megabalanus rosa

c

N

Arthropoda

Pycnogonida

Pantopoda

Phoxichilidium quadradentatum

a

N, may encyst in hydroids2

Arthropoda

Pycnogonida

Unknown

Unknown

c

1

Bryozoa

Gymnolaemata

Cheilostomatida

Bugula spp.

a, b, c

1

Bryozoa

Gymnolaemata

Cheilostomatida

Jellyella eburnea

a

1

Bryozoa

Gymnolaemata

Cheilostomatida

Jellyella tuberculata

a

1

Bryozoa

Gymnolaemata

Cheilostomatida

Jellyella/Membranipora

b,c

1

Bryozoa

Gymnolaemata

Cheilostomatida

Membranipora tenella

a

1

Bryozoa

Gymnolaemata

Ctenostomatida

Bowerbankia spp.

a

1

Bryozoa

Gymnolaemata

Ctenostomatida

Victorella spp.

a

N, may disperse through fragmentation of substrate3

Bryozoa

Stenolaemata

Cyclostomatida

Filicrisia spp.

a

N

Bryozoa

Stenolaemata

Cyclostomatida

Stomatopora spp.

a

N

Bryozoa

Stenolaemata

Cyclostomatida

Tubulipora spp.

a

1

Chordata

Perciformes

Actinopterygii

Abudefduf spp.(vaigiensis?)

b, c

N/A

Chordata

Perciformes

Actinopterygii

Canthidermis maculata

c

N/A

Chordata

Perciformes

Actinopterygii

Chirolophis spp.

c

N/A

Chordata

Perciformes

Actinopterygii

Coryphaena hippurus

b

N/A

Chordata

Perciformes

Actinopterygii

Elagatis bipinnulata

b

N/A

Chordata

Perciformes

Actinopterygii

Histrio histrio

c

N/A

Chordata

Perciformes

Actinopterygii

Kyphosus spp. (vaigiensis?)

b, c

N/A

Chordata

Perciformes

Actinopterygii

Meiacanthus spp.

c

N/A

Chordata

Perciformes

Actinopterygii

Seriola rivoliana

c

N/A

Chordata

Unknown

Unknown

Beige fish eggs

c

1

Chordata

Unknown

Unknown

Blue fish eggs

c

1

Chordata

Unknown

Unknown

Fish eggs

a,b

1

Ciliophora

Heterotrichea

Heterotrichida

Halofolliculina spp.

c

N on plastic, documented on wood4

Cnidaria

Anthozoa

Actinaria

Actiniidae

b

1

Cnidaria

Anthozoa

Actinaria

Anthopleura spp.

a.b

N, may disperse through detachment5

Cnidaria

Anthozoa

Actinaria

Calliactus sp.

c

7

Cnidaria

Anthozoa

Actinaria

Metridium spp.

a

N, may disperse through detachment5

Cnidaria

Anthozoa

Actinaria

Hormathiidae

c

1

Cnidaria

Anthozoa

Scleractinia

stony coral

b

1

Cnidaria

Hydrozoa

Leptothecata

Clytia gregaria

a

N, though 9 other Clytia species documented as rafting1

Cnidaria

Hydrozoa

Leptothecata

Obelia spp.

a

1

Cnidaria

Hydrozoa

Leptothecata

Plumularia setacea

a

1

Cnidaria

Hydrozoa

Unknown

Hydroid

b, c

1

Echinodermata

Ophiurodea

 

Ophiuroidea spp. 1

c

 

Echinodermata

Ophiurodea

 

Ophiuroidea spp. 2

c

 

Echinodermata

Ophiurodea

 

Ophiuroidea spp. 3

c

 

Foraminifera

Polythalamea

Rotallida

Planulina ornata

a

N

Mollusca

Bivalvia

Arcoida

Arcidae

c

N

Mollusca

Bivalvia

Myoida

Teredo spp.

c

1

Mollusca

Bivalvia

Myoida

Zirfaea spp. (pilsbryi?)

b

N

Mollusca

Bivalvia

Mytiloida

Mytilus galloprovincialis

a, c

1

Mollusca

Bivalvia

Ostreoida

Crassostrea gigas

b, c

1

Mollusca

Bivalvia

Pectinoida

Chlamys spp.

c

1

Mollusca

Bivalvia

Pteroida

Pinctada spp.

c

1

Mollusca

Bivalvia

Unknown

Lower valve of oyster

c

1

Mollusca

Gastropoda

Caenogastropoda

Litiopa melanostoma

c

1

Mollusca

Gastropoda

Littorinimorpha

Erronea spp.

c

N, may have widespread larval transport6

Mollusca

Gastropoda

Nudibranchia

Fiona pinnata

a, b, c

1

Mollusca

Gastropoda

Nudibranchia

Fiona pinnata eggs

a

1

Mollusca

Gastropoda

Pleurobranchomorpha

Berthella spp.

c

N

Mollusca

Gastropoda

Superfamily Pyramidelloidea

Odostomia (Evalea) tenuisculpta

a

N

Platyhelminthes

Rhabditophora

Polycladida

Rhabditophora

c

1

Platyhelminthes

Rhabditophora

Rhabdocoela

Rhabdocoela

c

1

Platyhelminthes

Turbellaria

Unknown

Flatworm

a, b

1

Platyhelminthes

Turbellaria

Unknown

Flatworm

b

1

Porifera

Calcarea

Leucosolenida

Sycon spp.

b, c

N

Porifera

Demospongiae

Halichondrida

Halichondria panacea

a

N

Years observed are a = Eastern Pacific 2009; b = Eastern Pacific 2011; c = Western Pacific 2012. Previously documented as rafting are N = Not listed as rafting in the scientific literature, N/A = fishes

1listed in Thiel and Gutow 2005b, 2 Lovely 2005, 3 Carter et al. 2010, 4 Matthews 1963, 5 Riemann-Zürneck 1998, 6 Emerson and Chaney 1995, 7 Bryan et al. 2012

Representatives of 11 phyla were found, with the most abundant phylum being the Arthropoda, followed by Mollusca and Cnidaria (Fig. 3a). The majority of these taxa were suspension feeders, though omnivores, grazers, and predators were also well represented (Fig. 3b). Slightly more taxa were mobile than were sessile (Fig. 3c). Of all 95 identified taxa, 25 (26 %) had not been previously found to occur in rafting assemblages (Fig. 3d, Table 1). No differences in the composition of phyla, feeding type, and mobile/sessile taxa were found between cruise years/locations (chi-square test, P > 0.5 for all tests).
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-014-2432-8/MediaObjects/227_2014_2432_Fig3_HTML.gif
Fig. 3

Characteristics of rafting taxa. a Phyla found on or around floating plastic debris. b Feeding types of rafting taxa. NA denotes eggs. c Mobile taxa versus sessile taxa. Fishes are excluded. d Taxa that have previously been documented to occur in rafting communities versus taxa that have not been documented. There was no statistical difference between cruise years/locations (chi-square test, P > 0.5 for all tests) for these analyses, so data from all cruises were combined in this figure. N = 95 taxa

We found a significant positive correlation between the size of the debris object and the number of taxa found on that object (Fig. 4a, Kendall’s tau, τ = 0.555, N = 242, P < 0.001). This correlation remained significant when the data were linearized through log transformation (Fig. 4b, linear regression, r2 = 0.169, F1.66 = 48.69, P < 0.001), as well as when the four largest items were removed (linear regression, r2 = 0.086, F1.66 = 22.45, P < 0.001). When cruises were examined separately (Fig. S1), there was a significant positive correlation between debris size and taxon richness in 2009 (Kendall’s tau, τ = 0.561, N = 208, P < 0.001) and 2011 (Kendall’s tau, τ = 0.650, N = 13, P = 0.003), but not in 2012 (Kendall’s tau, τ = 0.062, N = 21, P = 0.710). We did not find a relationship between distance offshore and number of taxa.
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-014-2432-8/MediaObjects/227_2014_2432_Fig4_HTML.gif
Fig. 4

Number of taxa versus surface area (m2) of debris object. a Semilog plot. Solid line is an exponential nonlinear least squares regression. Kendall’s tau, τ = 0.555, P < 0.001. b Log–log plot. Solid line is a linear regression (linear regression, r2 = 0.169, F1.66 = 48.69, P < 0.001). Symbols in both plots correspond to year and location of collection: eastern Pacific 2009 is shown in circles, eastern Pacific 2011 in triangles, and western Pacific 2012 in diamonds. Note that many of the symbols denoting small plastic objects overlap, making the sample size hard to discern visually. N = 242 debris objects

The eight most taxon-rich phyla also exhibited significant positive relationships between object size and number of taxa (Fig S2, Kendall’s tau P < 0.01 for all phyla). However, for the phyla that have few taxa (e.g., 1–3 taxa), these relationships are sensitive toward incidentally occurring individuals.

We noted a shallow parabolic shape, in both our overall taxa–area semi-log curve and for some of the phylum-specific curves, such as Arthropoda and Bryozoa. Higher numbers of taxa were found on medium-sized objects (approximately 1 × 10−2 m2) as compared to slightly larger objects (approximately 1 m2), though the largest objects (approximately 10 m2) retained the overall highest numbers of taxa.

Discussion

Composition of rafting assemblage

We found a diverse and widespread rafting assemblage inhabiting North Pacific plastic debris. The majority of taxa were known members of the rafting assemblage such as Lepas spp. barnacles and membraniporid bryozoans, but we documented 25 taxa that had not been previously found in rafting assemblages. Many of the previously undocumented taxa were from groups that are known to be prolific and successful rafters, such as the bryozoans, sponges, and peracarid crustaceans (Barnes 2002; Thiel and Gutow 2005b). We were surprised to find a small number of boring organisms rafting on pelagic plastic debris composed of foamed polystyrene. These included the bivalve Zirfaea, the shipworm Teredo, and a sphaeromatid isopod. While boring organisms are known to colonize rafts of biotic origin, such as wood and algae, they are relatively rare on plastic debris (Thiel and Gutow 2005b; Thiel and Haye 2006), although sphaeromatid isopods are known to burrow in polystyrene floats in coastal ecosystems (Davidson 2012).

Another organism of particular interest was the folliculinid ciliate (Halofolliculina spp.), found in abundance on some western Pacific plastic debris. These ciliates are pathogens that cause skeletal eroding band (SEB) disease in corals (Rodriguez et al. 2009). Though originally thought to be limited to the Indian Ocean and South Pacific, SEB disease was discovered in Caribbean corals in 2004 (Croquer et al. 2006) and in Hawaiian corals in 2010 (Palmer and Gates 2010). The mechanism behind the spread of SEB is not known (Croquer et al. 2006), but since the Hawaiian Islands are highly impacted by plastic debris collected by the North Pacific subtropical convergence zone (Dameron et al. 2007), it is possible that debris facilitated the dispersal of Halofolliculina to this area. Like many rafting substrates, plastic debris has the potential to disperse non-ciliate pathogens, such as viruses, but the role of debris as a disease fomite has been little studied (Maso et al. 2003; Pham et al. 2012).

Origin of rafting organisms

Many of the rafting taxa found are known invaders, but could have come either from their native range or from an area in which they are already established as a non-native species. These include the acorn barnacle Megabalanus rosa, native to Japan but an invasive species in Australia, and the mussel Mytilus galloprovincialis and acorn barnacle Amphibalanus amphitrite, which are invasive to the eastern Pacific (Fofonoff et al. 2012). Other taxa may have settled onto debris in coastal areas and been transported offshore (Astudillo et al. 2009), such as the bryozoan Victorella spp., which primarily occurs in estuarine waters (Carter et al. 2010).

Because most of the fauna present were either known members of the North Pacific rafting assemblage or widely distributed taxa, the source of the debris objects could not be determined from the associated assemblage. We know of no other reliable way to age or source plastic debris (though see the general analysis of net type in Jacobsen et al. 2010). Even debris with some identifying markings or text in a particular language cannot always be attributed to a country of origin due to the extent of international trade and the variety of household items used on ships.

The transport of invasive species on debris originating during the March 11, 2011, Tohoku Earthquake and subsequent tsunami event has received much attention recently, especially regarding objects such as docks that harbor entire communities of coastal organisms (Choong and Calder 2013; Gewin 2013). It is possible that some debris collected during the 2011 and 2012 expeditions originated from the Japan tsunami. However, we believe this is unlikely, since both expeditions occurred outside of the locations where high tsunami debris concentrations were predicted to occur at the time of the expedition (Lebreton and Borrero 2013).

Relationship between number of taxa and debris size

We found a greater number of taxa on larger debris items than on smaller items. A positive relationship between object size and taxa number has also been observed in algal rafts (Ingólfsson 1995; Hobday 2000; Clarkin et al. 2012), fish aggregation devices (Nelson 2003), and pumice (Bryan et al. 2012). The greater number of taxa on larger objects could be a stochastic effect. If individuals are randomly distributed over all floating objects available, then larger objects would receive more species. However, this relationship may also be driven by both physical and biological factors. Physically, larger objects are more likely to remain buoyant, even after developing a substantial rafting assemblage (Thiel and Gutow 2005a). Smaller objects with a substantial rafting assemblage may have become negatively buoyant and were therefore not sampled by this study or by other studies focusing on material at the sea surface. In addition, some objects were likely to have entered the water without attached biota (e.g., a toothbrush), while some objects were likely to have had some attached biota at the time they were lost (e.g., an eel trap). Larger items were primarily associated with fishing activities and may be more likely to have a higher number of taxa due to previous attachments.

Biological interactions, as predicted by island biogeography, may also be an important driver of the positive species–area relationship on pelagic plastic debris. As on islands, the rate of migration to large pieces of plastic debris may exceed the rate of extinction (Simberloff 1976). Higher rates of migration to larger objects could be driven by larval settlement. For example, due to their greater surface area, larger objects may be more likely to give off the appropriate cues for larval settlement (Rodriguez et al. 1993). Larger objects may also be easier for fishes to detect through visual and auditory cues (Dempster and Kingsford 2003). Furthermore, they are more stable at the sea surface, which is an important driver of diversity (Bravo et al. 2011; Bryan et al. 2012). Large items without spatial stability may contain low diversity, such as a meter-long cylindrical polystyrene buoy that “rolled” along the sea surface (H.S.C. pers. obs.). Lastly, successional stage may influence diversity. A study in the western Pacific found that diversity was highest during early stages of succession and then reduced by Lepas dominance (Tsikhon-Lukanina et al. 2001). The authors suggested that diversity may increase once more at later successional stages, which they define as being dominated by bivalves. However, floating debris in their study never reached this stage, perhaps due to lack of physical stability or limited larval supply. A study on algal rafts suggested that the separation and exchange of rafting materials may affect the successional progression, which may also apply to some marine debris (e.g., rope and net masses) but may not apply to solid plastic objects that are less likely to coalesce (Clarkin et al. 2012).

We have two potential explanations for the shallow parabolic shape of some of the taxa–area semi-log curves presented here: (1) it could be an artifact of different sampling methodologies for medium and large objects; (2) higher predator abundance on large objects suppresses prey diversity. We cannot rule out the parabola as a sampling artifact, because medium objects (e.g., plastic fragments) were preserved without subsampling, while large objects (e.g., buoys) had to be subsampled at sea. The difference in preservation strategy could have caused inconspicuous taxa on large objects to be undersampled. To address the sampling issue, future studies should consider subsampling larger items with non-selective methodology such as suction.

However, it is also possible that there is a threshold debris size beyond which fish and decapod predators suppress prey taxa diversity through direct predation and/or non-consumptive predator effects (Matassa and Trussell 2011). Some epipelagic fishes preferentially associate with fouled rafting objects as opposed to unfouled rafting objects, suggesting that some of these fishes may prey on rafting invertebrates (Nelson 2003, Thiel and Gutow 2005b), though other studies have not found significant predation on the fouling community (Ibrahim et al. 1996; Nelson 2003; Vassilopoulou et al. 2004). Most fishes observed in this study were associated with the three net and rope masses, with the exception of one juvenile triggerfish (Canthidermis maculata) on a bleach bottle and a school of sergeant majors (Abudefduf spp.) inside a plastic crate. The net-associated fishes exhibited all three types of behavior identified by Hirosaki (1960) for macroalgae-associated fishes: staying almost exclusively within the net mass, such as the sargassum frogfish (Histrio histrio); remaining underneath or around the mass in close association, such as the sergeants; and swimming around the mass without close association, such as the mahi mahi (Coryphaena hippurus). We also observed fish entangled in net masses that were still alive, recently dead, and partially decomposed. The conspicuously low abundance or absence of some species such as Lepas spp. and Halobates spp. eggs on net masses compared with isolated fragments may be partially attributed to these net-mass-associated fishes. For example, the forward half of a small fishing boat floating upright in the western Pacific had more than 50 associated individual fish, but very little attached biota (M.E. pers. obs.).

Plastic-associated rafting organisms may also be impacting the pelagic ecosystem by reworking the organic particle size spectrum through ingestion and egestion (Mook 1981). Suspension-feeding rafting organisms prey on a variety of particle sizes, from 3 to 5 μm for Mytilus mussels (Lesser et al. 1992), 10–20 μm for bryozoans (Pratt 2008), 20–125 μm for caprellid amphipods (Caine 1977), and 0.5 to more than 1 mm for lepadid barnacles and hydroids (Evans 1958; Boero et al. 2007; Goldstein and Goodwin 2013). This size range encompasses much of the non-microbial organic particle size spectrum of the oligotrophic North Pacific (Sheldon et al. 1972). Since organic particle size determines whether energy flows into the microbial loop or into the metazoan food web, large-scale alterations in particle size could substantially influence the species composition of the North Pacific subtropical gyre (Karl et al. 2001). Future research should also consider phytoplankton and microbial interactions with pelagic plastic macrodebris (Maso et al. 2003; Zettler et al. 2013; Carson et al. 2013).

Though the majority of plastic debris items in the North Pacific are small fragments (Goldstein et al. 2013), we found in this study that such particles carry few taxa, most of which are known subtropical rafters such as Jellyella or Membranipora bryozoans. We found the majority of displaced taxa on large items such as net balls, though the coral pathogen Halofolliculina spp. was found on medium-sized plastic fragments (0.03–0.1 m2). Species introductions from beached debris are most likely to occur on surfaces similar to the hard raft substrate (i.e., bedrock shorelines) and during large debris deposition events (i.e., storms). While debris removal operations are important to engaging the public on marine debris issues, they are unlikely to significantly reduce the chances of non-native species introductions from plastic debris. This is because coastal cleanups frequently occur on sandy beaches that are inherently inhospitable to rafting organisms and at pre-determined dates unlikely to coincide with major debris deposition events. While some rapid-response operations to large debris beachings, such as ships and docks originating from the 2011 Japanese tsunami event, may have been effective in preventing species introductions in limited circumstances, large-scale debris cleanup on rocky, remote shores is impracticable. Efforts to prevent debris input from both coastal sources (e.g., urban areas) and ocean sources (e.g., fishing debris) are likely the most efficient means of controlling debris-mediated species introductions.

Durable plastic “islands” are hosts to a diversity of coastal organisms in the pelagic environment, but the ecological role of plastic-associated rafting assemblages on the open ocean remains unclear. Whether or not the plastic rafts introduce new species to distant coastal regions, the consequences of these “misplaced” organisms to open-ocean ecosystems, especially in debris-accumulation zones, merits further study. However, any potential impacts of the debris-associated rafting community on coastal or pelagic ecosystems can be most effectively limited by an overall reduction in the quantity of plastic pollution introduced into the marine environment.

Acknowledgments

Funding for the 2009 SEAPLEX cruise was provided by University of California Ship Funds, Project Kaisei/Ocean Voyages Institute, AWIS-San Diego, and NSF IGERT Grant No. 0333444. The 2011 and 2012 expeditions were made possible through collaboration between the 5 Gyres Institute, Algalita Marine Research Institute, and the University of Hawaii, Hilo. The Sea Dragon was made available by Pangaea Explorations. M.C.G. was supported by NSF GK-12 Grant No. 0841407 and donations from Jim and Kris McMillan, Jeffrey and Marcy Krinsk, Lyn and Norman Lear, Ellis Wyer, and the Petersen Charitable Foundation. Funding for H.S.C. was provided by the Will J. Reid Foundation through a grant to K. McDermid. Laboratory supplies, support, and some analytical equipment for the 2009 samples were supplied by the SIO Pelagic Invertebrate Collection and the California Current Ecosystem LTER site supported by NSF. Many thanks to the captain and crew of the R/V New Horizon and Sea Dragon for their assistance in debris collection and many other logistics. We are grateful to H. Cha, M. DeMaintenon, M. Forrest, E. Moore, L. Sala, A. Townsend, and J. Winston for their assistance with taxonomy and to S. Strutt for her work in the laboratory. Comments from L. Gutow, M.R. Landry, M.D. Ohman, M. Thiel, and an anonymous reviewer significantly improved this manuscript.

Supplementary material

227_2014_2432_MOESM1_ESM.pdf (304 kb)
Supplementary material 1 (PDF 304 kb)
227_2014_2432_MOESM2_ESM.xlsx (69 kb)
Supplementary material 2 (XLSX 69 kb)

Copyright information

© Springer-Verlag Berlin Heidelberg 2014