Port-by-port accumulations and dispersal of hull fouling invertebrates between the Mediterranean Sea , the Atlantic Ocean and the Pacific Ocean

The R/V Oceanus completed a 9,789 km, 28 day passage from Woods Hole, Massachusetts, in the Atlantic Ocean, through the Panama Canal to Yaquina Bay, Oregon, in the Pacific Ocean on 21 February 2012. The Oceanus had previously operated in the Mediterranean Sea and Atlantic Ocean (including the Caribbean Sea). We document the sequential acquisition of the barnacles Balanus trigonus and Amphibalanus venustus and the oyster Ostrea equestris on the Oceanus on its high and low latitude transoceanic, intra-oceanic, and interoceanic travels before she was surveyed in Yaquina Bay. The close correspondence between hull fouling accumulations and the detailed two year Oceanus working history reveals B. trigonus settlement occurred in every tropical port visited by the Oceanus, that some populations survived through two of three Woods Hole winters, and that some of these populations passed through the freshwater Panama Canal. These results suggest that marine hull-fouling species are continuously transported globally between most ports of call by most ship passages.


Introduction
The port-by-port development of fouling communities on vessels over their cruise tracks is well known but rarely studied (Darwin 1854;Pilsbry 1916;Kerckhof et al. 2010;Carlton 2011;Carlton et al. 2011) and was not quantitatively or experimentally examined relative to a detailed vessel itinerary until the 1980s (Carlton and Hodder 1995).Kerckhof et al. (2010) and Floerl and Coutts (2009) proposed that more invasions from hull fouling species are to be expected because global shipping has increased, and ships are faster and increasing their numbers of ports of call.Kerckhof et al. (2010) considered also the increased potential of dispersal on ships that operate for long periods of time in local areas and then periodically sail to distant areas.
Major intra-and inter-ocean salinity and temperature variation could limit the scope of some of Kerckhof et al.'s (2010) and Floerl and Coutts' (2009) predictions.The freshwater Gatun Lake in the Panama Canal, for instance, is commonly assumed to limit dispersal of marine organisms between the Atlantic and Pacific oceans.However, marine species survival through the canal on the hulls of ships has been reported multiple times (Cohen 2006), and passage through the Canal in salt water ballast water avoids exposure to Gatun Lake.Bishop (1947) reported that three species of barnacles, Austrominius modestus (Darwin, 1854) (= Elminius modestus), Amphibalanus amphitrite (Darwin, 1854) (=Balanus amphitrite) and Amphibalanus improvisus (Darwin, 1854) (= Balanus improvisus) arrived alive in Liverpool, England, after a 30-day voyage from Australasia, via the Panama Canal.Menzies (1968) found high survival through Gatun Lake of marine intertidal snails, crabs, and barnacles and subtidal isopods suspended in a cheesecloth bag that was towed from a ship transiting the Canal.Davidson et al. (2008) found surviving barnacles, bryozoans, and isopods, among other hull fouling species, on ships that passed through the Panama Canal in transit from a low salinity area of San Francisco Bay, California, to Brownsville, Texas.
Between 1975 and 2012, the Atlantic Ocean was the home port and research arena of the 54 m long R/V Oceanus of the Woods Hole Oceanographic Institution, Woods Hole, Massachusetts.The Oceanus was drydocked and cleaned in Jacksonville, Florida, between February and March 2009 and then operated between 2009 and 2011 in marine waters of the Mediterranean Sea, Atlantic Ocean (including the Caribbean Sea) before she was transferred to the Pacific Ocean via the Panama Canal in 2012.The Oceanus took 28 days (25 January to 21 February 2012) to travel the 9,789 km from Woods Hole to Yaquina Bay, Oregon.The passage included an approximately 6 hour (4 to 5 February) transit through the Panama Canal (Figure 1).
We surveyed the hull of the R/V Oceanus on its arrival in Yaquina Bay to partially test Cohen's (2006), Kerckhof et al.'s (2010), and Carlton's (2009) proposals of persistent dispersal and survival by hull fouling organisms on modern ships moving between local and transoceanic ports across major ocean barriers.The sister ship of the Oceanus, the R/V Wecoma, was immediately adjacent to the Oceanus at this time and had been continuously moored at the dock for several months and was covered by a local species community that could be compared with the Oceanus species community.We therefore surveyed the hull of the Wecoma simultaneously with the Oceanus.

Methods
Oregon State University (OSU) and Oregon Coast Aquarium (OCA) scientific divers surveyed the hull of the Oceanus by direct sampling and by underwater videos on 23 February 2012.All accessible macro-fouling invertebrate assemblages were sampled.Samples were scraped into plastic bags that were then transferred into seawater baths at the OSU Hatfield Marine Science Center (HMSC) facilities.Samples were sorted and preserved within 24 hours.Invertebrates were separated from these samples, photographed, identified to the lowest possible taxonomic level, and preserved in 95% ethanol.The specimens are deposited at the Natural History Museum of Los Angeles County, Los Angeles, California.
We compared the size structures of species populations among sequential settlement layers contained in the fouling communities of both vessels.We noted the order of occurrence of species among these fouling layers and the cohort sizedistributions within these species populations.We inferred the order of settlement and relative rates of individual growth on the Oceanus by comparing population size-frequency modes.We then back calculated the timing of their settlement relative to the timing of Oceanus voyages and home port layovers.The timing and destinations of the Oceanus cruises since its last drydocking and cleaning (in February-March 2009) and its passage from Wood Hole, Massachusetts, to Yaquina Bay, Oregon, were determined from the online Oceanus cruise log (http://www.whoi.edu/main/ships/schedules) and from additional information provided by Oceanus Captain Jeff Crews (pers.com.).
We assumed thermal tolerances of barnacles on the Oceanus to be similar to other populations of the same species (e.g., Werner 1967) in order to assess their survival in Woods Hole sea water temperatures.We obtained Woods Hole 12 m depth seawater temperature records from the Martha's Vineyard Coastal Observatory (MVCO 2012) database.Missing temperature data were interpolated between the nearest beginning and ending dates of available data.
Barnacles and oysters containing fresh tissue were classified as alive at time of sampling.Conversely, these invertebrates were classified as dead at time of sampling when they lacked fresh tissue or when their internal test or shell walls were overgrown by other fouling species.Live barnacle identifications were from the coloration and shapes of their shells and their opercular valve characters (Zullo 1979;Newman 2007).We identified dead barnacles from the close match of their shell morphologies with the live-barnacle shell morphologies.Order of settlement on the Oceanus was determined from overgrowth positions on previously settled layers and by the temperate or tropical origins of species forming the layers.Invertebrates attached directly to the hull were readily identified from their uniformly flat bases or adhering antifouling paint.We assumed that barnacle and oyster diameters and lengths increase with age and inferred relative ages of these taxa from base diameters or shell lengths measured to the nearest 0.1 mm.Lateral crowding likely to bias conspecific barnacle size estimates from length or diameter dimensions was uncommon in these samples.
We used the distinction of size modes to test whether barnacle recruitment was constant or irregular and distinguished population sizefrequency modes using the FAO FiSAT_II software (FAO 2005).FiSAT_II decomposes composite size distributions into their component cohorts (means, standard deviations and population sizes of modes) and subjectively identifies and "links" the mean sizes of the most likely common cohorts.The FiSAT_II procedure uses the Bhattacharya method for distinguishing particular size modes and the NORMSEP program therein decomposes composite size modes by the simplex algorithim to a maximim fit of estimated and observed total abundances across entire size distributions.Adjacent size modes are accepted as significantly different in the FiSAT_II procedure when their separation indices (S.I.) are greater than two (FAO 2005).We tested the significance of estimated size frequencies (the sum of all modes) by their correlations with the observed size frequency distributions.

Oyster settlement dates
We back calculated the probable oyster settlement dates from estimated growth rates derived from two Atlantic populations of the crested oyster Ostrea equestris Say, 1834 (averaging 18 and 20 mm in length) that had settled onto buoys that were exposed for 12 months (Galtsoff and Merrill 1962).We assumed these populations began settlement in November when the buoys were placed out and increased in length by logistic growth and estimated their high and low daily growth increments, r, over one year as follows: r = ln(Final length/Initial length)/365 1).
We used our high and low estimated daily growth increments: r high = 0.0063 d -1 and; r low = 0.0006 d

Results
Diver accounts and videos indicated that less than 1% of the Oceanus' hull surface was covered by hull-fouling invertebrates.Most of these fouling organisms occurred where blocks had supported the vessel during its 2009 dry docking.Lesser abundances of fouling organisms were found in recesses of the keel and rudder.Nearly 100% of the Wecoma's hull was covered by hull-fouling invertebrates.

Species
Fouling species on the Oceanus included the temperate and subtropical northwest Atlantic barnacle Amphibalanus venustus (Darwin, 1854) (Figures 2A and 2D), the warm temperate and tropical Pacific barnacle Balanus trigonus Darwin, 1854 (Figures 3A and 3E), and the warm temperate and tropical eastern Atlantic species the crested oyster Ostrea equestris (not shown).
The Atlantic bryozoan Biflustra arborescens (Canu and Bassler, 1928) was also present, all of which were dead.Mud tubes were common among the Oceanus barnacle samples but were empty.Additionally, resident Yaquina Bay isopods, specifically the introduced Sphaeroma quoyanum H. Milne Edwards, 1840 and the native Gnorimosphaeroma oregonensis (Dana, 1853), were found on the Oceanus hull, presumably derived from adjacent marine structures in the two days between the vessel's arrival and its sampling.Similar densities of these isopods, in the same size ranges as were found on the Oceanus, were found in the simultaneous samples of the Wecoma.These two isopods commonly disperse at night (JWC, personal observations), and thus are likely to have settled onto both vessels on the previous two nights.
We found dense settlements of the native barnacles Balanus glandula Darwin, 1854 and Balanus crenatus Bruguière, 1989 on the hull of the Wecoma.Minimum diameters among all specimens of these two species were greater than 1.0 mm and we found no barnacle recruits less than 48 hours old.Similarly, we did not find newly-settled recruits of either of these two native species on the Oceanus.

Barnacle settlement
Live and dead B. trigonus and A. venustus occurred in 5 layers on the Oceanus (Table 1).Base diameters of all live barnacles were less than 15.1 mm (Figure 4).The dead A. venustus were more frequent than live, occurred in all of the largest size classes and, overall, were predominantly smaller than the live A. venustus (Figure 4A).The five major barnacle layers (Figure 4B and 4E) consisted of A. venustus in layers 1, 3, and 5 (Figure 4B) and B. trigonus in layers 2 and 4 (Figure 4E).Dead B. trigonus that could be classified relative to settlement layers were between A. venustus layers 1 and 3 or under layer 5. Live B. trigonus that could be classified relative to settlement layers were exclusively between A. venustus layers 3 and 5. Thus, dead B. trigonus (Figure 4D) occurred almost exclusively in layer 2 (Figure 4E) while all live B. trigonus (Figure 4D) occurred in layer 4 (Figure 4E).
A single A. venustus size class was apparent among the total live populations (Figure 4A The close correlation between proportions of live A. venustus and base diameter (Figure 4C) indicates predominantly size-dependent mortality among the cohorts over time.The similar overall frequencies of A. venustus diameters among layers 1, 3, and 5 and among live and dead populations, however, indicate mixed classifications or other complications occurred with this species.These complications can result from variable individual growth, uncertain classifications of layers, or mixing of layers by simultaneous settlement on multiple layers.These challenges precluded further analyses of A. venustus growth.
Differences in frequency modes of dead and live B. trigonus base diameters (Table 2) were nearly perfectly correlated with the timing of the Oceanus visits to diverse tropical ports of call (Figure 5, Table 3).These size modes thus appear to be separate cohorts.The 25 January 2011 intercept of dead B. trigonus (Figure 5

Oyster settlement
All of the 11 Ostrea equestris recovered from the Oceanus were alive (Table 1), similar in length (mean length = 8.7 mm S.D. = 2.2) and prereproductive.They were attached directly to the Oceanus hull, and overlapped B. trigonus layers 2 and 4 and A. venustus layers 1, 3, and possibly 5 (Table 1).A 1.0 mm A. venustus that settled onto a 9.5 mm O. equestris places that particular oyster in layer 4. Thus, all of the similar sized O. equestris are likely to have settled on layer 4.Moreover, the possible layer 5 oyster we found A. venustus on (Table 1) was likely a misclassified individual from layer 3.
The mid date between the earliest and latest likely O. equestris settlement is 4 July 2011 and the average of the early and late settlement estimated from growth time (equation 2) to the 8.7 mm average O. equestris length is 30 June  2011.The similar sized O. equestris (Table 1) are likely to have settled simultaneously from a single low latitude port along with B. trigonus.
Our earliest growth estimated O. equestris settlement (equation 3) E Set Day, 18 May 2011, was not early enough to overlap the three day Oceanus visit to Bridgetown, Barbados that began on 13 May 2011 (Table 3) but does overlap its two visits to Morehead City, North Carolina followed by St. Georges, Bermuda (Figure 5,  3), and the second visit to Morehead City (Figure 5) began 18 June 2011 and lasted 3 days.Thus, O. equestris was unlikely to have settled in the Barbados and was most likely to have settled either in Morehead City, along with B. trigonus live cohort 2 or in Bermuda, along with B. trigonus live cohort 3 (Table 3, Figure 5), in deposition layer 4 (Table 1).

Discussion
We are unaware of previous detailed quantitative analyses of recruitment, growth, survival, and transport of hull fouling species among multiple ports over time and between oceans.The Oceanus was maintained at the "clean" end of the vessel fouling spectrum (Davis et al. 2008) but nevertheless accumulated species in every port and carried nonindigenous species to most subsequent ports of call.All unprotected immersed surfaces of every vessel in every marine port are thus likely to be vulnerable to accumulations of settling marine organisms.These organisms are readily dispersed to every following port of call, even if they pass through the freshwaters of the Panama Canal and even after relatively long intervening residences in seemingly hostile climates.Despite ever increasing prevention efforts, these findings corroborate Carlton's (2011) and Kerckhof et al.'s (2010) proposals that port-toport and ocean-to-ocean dispersal of hull fouling organisms, that has existed for centuries and millennia (Carlton 2009), continues on modern ships.Passages of Balanus trigonus on ships around Cape Horn or the Cape of Good Hope into the Atlantic Ocean must have been commonplace for over 300 years previous to its 1800s discovery in the Atlantic Ocean, and previous to the openings of the Suez or Panama Canals (Carlton et al. 2011).Continuing transport of B. trigonus in both directions through the Panama Canal, around Cape Horn and Cape of Good Hope, and among all tropical and subtropical marine ports ensures that a constant supply of propagules is available to maintain or expand its range, which now will be further facilitated by climate warming at higher latitudes.Darwin (1854) described a sequence of barnacle recruitment to the hull of a vessel plying the South Atlantic Ocean between Africa and South America.In Darwin's sequence, Megabalanus tintinnabulum (Linnaeus, 1758) (layer 1) settled on the ship in Namibia on the West African coast; then, in Patagonia, Austromegabalanus psittacus (Molina, 1788) (layer 2) settled on the African-acquired M. tintinnabulum.On the return voyage to England, a third layer consisting of M. tintinnabulum settled on the A. psittacus.Pilsbry (1916) described barnacle recruitment on a whaling vessel in 1879 during a 6 month voyage in the Caribbean and its return to Cape Cod in New England.In the West Indies the vessel was colonized by Balanus trigonus (layer 1), upon which Megabalanus tintinnabulum (layer 2) then settled.The tropical barnacle Newmanella radiata (Bruguière, 1789) (layer 3) then settled on the Megabalanus before the vessel sailed north to New England.Upon arrival in Cape Cod, Amphibalanus eburneus (Gould, 1841) (layer 4) settled on the earlier (larger) barnacles.Recently, Kerckhof et al. (2010) described the acquisition of four barnacle species by a vessel spending several months along the west coast of Africa between the Gulf of Guinea and Sierra Leone.The barnacle Striatobalanus amaryllis (Darwin, 1854), represented by several cohorts, was found "intermingled" with Amphibalanus reticulatus (Utinomi, 1967).Megabalanus tintinnabulum then overgrew both species.While Chthamalus dentatus Krause, 1848 was also present, Kerckhof et al. (2010) did not resolve an exact number of settlement sequences.
The extended and dense settlement of Balanus glandula or Balanus crenatus during the Wecoma's extended moorage in Yaquina Bay precluded opportunities for detailed size analyses.The absence of 1.0 mm or less (newly settled) barnacles in the Wecoma samples however, precluded Yaquina Bay as a source of small, difficult to identify barnacle settlement on the Oceanus, in the 48 hours prior to sampling.Misidentified rare species among dead barnacle populations we examined are therefore unlikely sources of bias in these analyses.Thus, all barnacles of the Oceanus, even those 1.0 mm diameter or smaller, apparently settled prior to the Oceanus arrival in Oregon.
We discovered at least 9 Balanus trigonus settlement events during the Oceanus voyages in the Mediterranean and the western Atlantic theaters of operation.These 9 B. trigonus settlement events were in two settlement layers that were interspersed between 3 layers of Amphibalanus venustus.Due to their overgrowth by uniformly smaller live B. trigonus, the greater than 15 mm diameter population of dead B. trigonus on the Oceanus were thus likely to have lived the longest and were also the first to settle.
A nearly perfect correlation exists between the 5 dead B. trigonus size modes and the timing of Oceanus visits to tropical ports and a major mortality event between 2009 and 2011.The ports of call began with Piraeus, Greece, and were followed by western Atlantic, Gulf of Mexico, and Caribbean Sea ports.Growth rates of these barnacles were approximated by the slope of cohort base diameter over time.All of the tropical ports in question are in regions where B. trigonus occur.Moreover, these results indicate that geography, time, and extreme temperatures were the dominant variables over many other factors (including predation and salinity variations) likely to influence the survival and growth of the barnacles on the Oceanus.
The intercept of the dead B. trigonus populations was 25 January 2011, when Woods Hole seawater temperatures were at their minimum.) and 8 mm diameters in 25 days (~0.3 mm d -1 ) in the New Zealand summer (Ayling 1976).Balanus trigonus can produce multiple broods per year with peak reproduction in spring and fall (Werner 1967;Williams et al. 1984).Maximum B. trigonus growth in southern Florida is in the first 10 weeks after settlement where they require 14 weeks to reach 10 mm base diameters (~0.1 mm d -1 ) in January through April and 6-8 weeks in June through July (~0.2 mm d -1 ) (Werner 1967).Newly settled B. trigonus reach 6 mm diameters on fouling plates in the Colombian Caribbean in 20 weeks (~0.04 mm d -1 ) (Garcia 1977).Thus, the growth rates of the live and the dead B. trigonus populations on the Oceanus were relatively low.
The linear B. trigonus base diameter changes over time were not expected.We expected instead, given that base diameter is only two of the three dimensions of size, that B. trigonus diameter increases would decline approximately to the square root of time.The three dimensions (base diameter + height) of these barnacles appeared proportional and, at their relatively low densities, they were not altered greatly by crowding.Werner (1967, figures 5 and 6) reported overall declining average rates of diameter increase with time among Miami Beach, FL populations of B. trigonus, as we expected.However, base diameters of Werner's (1967) 4 m deep and slowest growing B. trigonus also increased linearly (Diameter = 0.046 mm d -1 + 5.459, r² = 0.98, df = 11) over the last 156 days of his 185 day study.
We also did not expect growth rates of the first 5 (dead) B. trigonus cohorts to be different from the growth rates of the last 4 (living) cohorts.The different growth rates of the two populations indicate that different conditions occurred during their growth periods.The fouling communities on the widely ranging Oceanus were unlikely to have experienced uniform environmental conditions.Moreover, the changing environmental conditions that characterized the Oceanus voyages are expected to result in different overall growth rates of older and younger cohorts and, thus, non-linear or variable growth rather than the closely linear growth that we observed.
The native Atlantic barnacle Amphibalanus venustus appears to have settled on the Oceanus while she was in Cape Cod and also when she was further south along the Atlantic coast.Amphibalanus venustus settled on the hull, on dead and live Balanus trigonus and on Ostrea equestris.A broad size range of A. venustus also survived the transit through the Panama Canal.Amphibalanus venustus settlement on the Oceanus is likely to have continued during most of its extended stays in Woods Hole between 2009 and 2012.Lang and Ackenhusen-Johns (1981) reported abundant A. venustus larvae in lower Narragansett Bay, Rhode Island (near Woods Hole) from May through December when water temperatures were consistently above 5 o C. As might be expected, A. venustus mortality did not increase in low seawater temperatures, including the 25 January 2011 low that was fatal to B. trigonus.

Potential cryptic distributions of Amphibalinine barnacles
A re-examination of Amphibalanus amphitrite and A. improvisus populations along the Pacific coasts of North America and South America is warranted to test for the cryptic presence of A. venustus in these regions (Carlton 2011).Amphibalanus venustus has been introduced to Japan (Otani 2006(Otani , 2012) ) and to South Africa (Mead et al. 2011), and it may exist elsewhere in the Pacific Ocean where it may be confused with other amphibalinine species.

Risk of Oceanus barnacles and oysters invading the Oregon coast
There is no indication that the warm-water barnacle Balanus trigonus and oyster Ostrea equestris found on the Oceanus were likely to colonize the cold waters of the Northeast Pacific north of California.A commercial diver nevertheless removed any barnacles remaining on the Oceanus after the survey (Jeff Crews, personal communication).Balanus trigonus occurs north to Monterey Bay (37 o N latitude) (Newman 2007) on the Pacific coast of North America.However, B. trigonus is carried almost everywhere in the world by ships that have previously visited low latitude ports (Carlton et al. 2011, Table 3).Ostrea equestris does not occur in the cooler waters north of Cape Hatteras, North Carolina (Galtsoff and Merrill 1962;Mikkelsen and Bieler 2008) or in river dominated estuaries.The Oregon coast water temperatures and the river dominated Yaquina Bay are therefore unlikely to be sufficiently warm for marine for O. equestris reproduction.In contrast, Amphibalanus venustus is found north to Cape Cod in New England (Zullo 1979) and has previously been found in Japan (Otani 2006).This barnacle may be capable of colonizing the cold water bays of the Pacific Northwest, where it could be readily confused with the established and earlier-introduced barnacle Amphibalanus improvisus.
) coincides with the lowest and most lethal sea water temperatures of the 2010-2011 winter (0.5 o C).These were also the lowest temperatures that occurred between 2009 through 2011.The successively larger base diameter cohorts of B. trigonus were thus, as expected, the first to settle on the Oceanus.However, progressively smaller B. trigonus on the Oceanus settled in precisely linear correlations with the timing of Oceanus ports of call, time previous to the coldest water temperatures in Woods Hole (25 January 2011), and to our 2012 survey (Figure 5 regression lines).

Figure 5 .
Figure 5. Degrees north latitude positions of the Oceanus (thick black line), sea surface temperature at Woods Hole, MA ( o C) (red line), critical 5 o C minimum B. trigonus temperature range (dashed horizontal line), average base diameter (mm) of dead cohorts 1-5 (open circles) and live cohorts 1-4 (solid blue circles) of Balanus trigonus and respective regression equations with day*, departure date of Oceanus from Woods Hole (broken circle) and hull fouling survey date (crosshatch).* Note: Day = days previous to 2/23/2012 survey adjusted to Excel day zero [1/1/1900]) for this graph.
Thus, pre-2010 to fall 2011 settlements of B. trigonus succumbed to Woods Hole winter temperatures but 2011 and later settlements survived shorter exposures to Cape Cod winters.The B. trigonus that settled in Piraeus, Greece, around 26 July 2009 and Jacksonville, Florida, around 27 September 2009 survived the relatively short 2009-2010 winter but died around 25 January 2011 with the three following cohorts in the coldest period of the long 2010-2011 winter.These results are consistent with previous measures of B. trigonus temperature tolerances.Werner (1967) observed 50% mortality of B. trigonus in 0 o C seawater at 9 days and found some survival after 18 days.The average and minimum December through February Woods Hole area seawater temperatures of 2011-2012 (previous to the 25 January departure of the Oceanus) were respectively, 4.8 and 3.2 o C, higher than in the December through February period of 2010-2011 The lowest 2011-2012 temperature previous to the Oceanus' departure was 3.6 o C. The small (< 10 cm diameter) live B. trigonus appear to have settled on the Oceanus while she was in the Caribbean in 2011 but survived for 45 days below 10 o C in the milder (> 5 o C) 2011-2012 Woods Hole winter temperatures before the 25 January 2012 Oceanus departure.The largest three live B. trigonus cohorts that reached Yaquina Bay had settled after the 2010-2011 winter.These live cohorts had also survived a short stay in the relatively mild 2011-2012 Woods Hole winter until the Oceanus' departure and passage through the freshwater Lake Gatun in the Panama Canal and arrival in Yaquina Bay on 23 February 2012.These results are corroborated by Bishop's (1947) previous observations of B. trigonus passing through the Canal in the opposite direction on the hull of a vessel sailing from Australia and New Zealand to Liverpool in 1946.The smallest B. trigonus cohort recovered from the Oceanus is thus likely to have settled in the Pacific Ocean in Panama City, Panama rather than in the Gulf of Mexico or Atlantic Ocean.Since March 2009, the Oceanus was either (1) berthed in Woods Hole where high plankton densities are expected, (2) in the low production open sea where low plankton densities and poor barnacle growth conditions are expected, or briefly, (3) in tropical ports.In coastal waters, B. trigonus can reach reproductive size at 4 mm base diameter in as little as 3 weeks (~0.2 mm d - 1

Table 1 .
Live and dead barnacles Amphibalanus venustus and Balanus trigonus and the oyster Ostrea equestris observed in layers on the Oceanus beginning with layer 1 on the hull and sequentially superimposed layers (L2-L5).

Table 3 )
. Our estimated latest O. equestris L Set Day, 19 August 2011, occurred when the Oceanus was in Woods Hole (which is not possible because O. equestris does not occur north of Virginia; Mikkelsen and Bieler 2008), and previous only to its 13 October 2011 visit to St. Georges, Bermuda, (where O. equestris does occur; Mikkelsen and