First record of the non-native light bulb tunicate Clavelina lepadiformis ( Müller , 1776 ) in the northwest Atlantic

We report the first record of the colonial tunicate Clavelina lepadiformis (Müller, 1776) in the northwest Atlantic. Populations were found along the eastern Connecticut shoreline in October 2009. At one site C. lepadiformis had a mean percent cover of 19.95% (±4.16 S.E.). A regional survey suggests that the invasion is relatively localized. Genetic analysis confirms our morphological identification and places the introduced population in the previously described ‘Atlantic clade’. While it appears Clavelina lepadiformis is currently in the incipient stage of introduction in eastern Connecticut waters, its spread to other areas in the region could lead to competition with resident members of shallow water epifaunal assemblages and shellfish species.

Clavelina lepadiformis (Müller, 1776) is a distinctive and conspicuous colonial tunicate, commonly called the light bulb tunicate (Figure 1).Its native range was initially reported to be from the Shetland Islands and Bergen, Norway in the north to the Bay of Biscay, the Mediterranean, and the Adriatic in the south (Berrill 1951).During the 1990s, C. lepadiformis was reported in the Azores and Madeira, Portugal (Wirtz and Martins 1993;Wirtz 1998).During the 2000s, C. lepadiformis invaded Knysna Estuary and Port Elizabeth, South Africa (Primo and Vazquez 2004;Robinson et al. 2005).
We report the first record of Clavelina lepadiformis in the northwest Atlantic.Relatively dense but localized populations were found in southeastern CT, USA: specifically, Stonington Harbor (October 16, 2009, sites 1 and 2; Annex 1) and the lower Thames River estuary (November 11, 2009, sites 15-16; Annex 1).Elsewhere colonies are reported to be limited to shallow littoral habitats (<50m) occupying natural or artificial hard substrates (e.g., rocky outcrops, wooden docks) and are known to dominate disturbed habitats (Naranjo et al. 1996;Picton and Morrow 2007).We found them to occur only on vertical man-made structures (e.g., breakwaters, pilings) in relatively shallow water depths (to 3 m).The species can tolerate a relatively broad range of salinities (14 -35 psu;Millar 1971).At both sites where we found C. lepadiformis, salinities were 29.5 psu, although near bottom salinities can be ~ 20 psu in the lower Thames River during the spring.Colonies typically consisted of a rosette of zooids showing connection only by a common basal test or stolons.Colonies were composed of a few, tens or hundreds of zooids, and fully developed zooids were 5-6 cm long.The thorax of C. lepadiformis is reported to be clear except for white, yellow, or pink bands around the oral siphon and along the dorsal lamina (Berrill 1951).All individuals found at Stonington Harbor and the Thames River had white pigmentation around the oral siphon (Figure 1).
As with all colonial tunicates, C. lepadiformis reproduces sexually, and brooding larvae reside at the base of the zooid atrial chamber.Once released, the free swimming period of the tadpoles is ~3 hr.Following settlement, formation of oozooids is completed in ~2 to 3 days (Berrill 1951).Colonies collected from Stonington Harbor contained numerous brooded larvae and embryos.Following its discovery, we conducted preliminary surveys to estimate percent cover of C. lepadiformis on the dock where it was initially found in Stonington Harbor (site 2, Annex 1) as well as qualitatively examining other natural and artificial surfaces of the harbor and adjacent coves and embayments within ~ 16 km to assess its current distribution (Figure 2).Two 20 m transects were run at 1 m and 2 m below MLW along the dock using SCUBA.Underwater photographs were taken at random locations along each transect (n = 8) and percent cover of C. lepadiformis was estimated using 100 grid counts placed over the images.Colonies occupied on average 19.95% ±4.16 S.E.(range =36.59% to 6.12%.) of space on the dock.Other invasive ascidians can have percent cover on this order (Osman and Whitlatch 2004;Djikstra et al. 2007).Our search for the species at nearby docks and other structures in the harbor revealed the species was only present on the pier immediately adjacent to the primary dock (within 65 m) but at much lower densities (5 colonies; ~< 1% cover).In order to confirm our morphological identification, a 586 base pair fragment of the mitochondrial gene CO1was isolated from two colonies of C. lepadiformis, one from either side of the pier in Stonington Harbor where the species was initially found (site 2, Annex 1).DNA was isolated from the branchial sacs of individual zooids using a CTAB based method, PCR amplified with tunicate-specific primers, and sequenced from plasmids (Stefaniak et al. 2009 for further details and primer sequences).The resultant sequences were aligned in Clustal X (v1.83,Chenna et al. 2003) with sixteen sequences from nine species of Clavelina obtained from GenBank as well as a sequence from Didemnum vexillum as an outgroup (Annex 2).The final alignment was 369 bases in length with no gaps.A maximum likelihood analysis was implemented in PAUP* (Swofford 2003) using the GTR+G substitution model (e.g.Lavane et al. 1984).The substitution model was chosen based on the AIC scores of several models, and the model parameters values  1).Branch lengths on the consensus tree (Figure 3) were optimized in PAUP* using the same model.
Bootstrap support was weak over much of the tree (Figure 3), most likely due to the short length of the alignment, however, samples from Stonington Harbor morphologically identified as C. lepadiformis group strongly (94% support) with other samples of C. lepadiformis and weakly (52% support) with samples identified as belonging to the Atlantic clade of the species by Turon et al. (2003).The two samples from Connecticut were identical to haplotype IX from Turon et al. (2003) which was the most common haplotype sampled in that study and was found throughout the native range of C. lepadiformis as well as in the Azores which is considered to be a non-native location (Wirtz and Martins 1993;Turon et al. 2003).
It is likely that the introduction of C. lepadiformis to the northwest Atlantic was human-mediated given its short non-feeding larval life span and the past history of humanmediated dispersal of other tunicate species (Lambert and Lambert 1998;Lambert 2001).Turon et al. (2003) found only colonies of C. lepadiformis belonging to the Atlantic clade exclusively inside harbors in the Mediterranean and not on sites with natural substrates suggesting that the Mediterranean harbor populations are the result of ship-mediated invasion from the Atlantic.The introduction likely resulted from hull fouling, fouling of sea chests, or (less likely) larvae transported in ballast water (Carlton and Geller 1993;Coutts and Dodgshun 2007).It is presently uncertain whether populations in Stonington Harbor and the Thames River were established from individuals from European populations or are the result of a secondary invasion within the NW Atlantic region.The Thames River has a variety of maritime operations including commercial shipping operating from the New London State Pier, private boating, fishing vessels, and berthing for cruise ships.The Thames River is also home to a U.S. Coast Guard station and a U.S. Navy submarine base.Stonington Harbor has a commercial fishing fleet and private recreational boating.Locke (2009) predicted C. lepadiformis as one of the most likely species to become established in Atlantic Canada, because of its broad native and invaded distributional range.While potential ecologic and economic impacts of this new invader are presently unknown, the species is known to form relatively dense aggregations (de Caralt et al. 2002 per. obs.) and may have the potential of over-growing and out competing resident fouling species and economically important shellfish.As the distribution of C. lepadiformis appears to be localized in southeastern Connecticut, we recommend an increase in surveying efforts along the southern New England coastline to better understand its current distribution and suggest that the species may be a good candidate for undertaking an immediate eradication program in order to prevent its spread.

Figure 2 .
Figure 2. Map of eastern Connecticut shoreline.Closed circles represent sites where C. lepadiformis is present and open circles where it is absent.Circles correspond to survey sites listed in Annex 1.Not all surveyed sites are shown.
No colonies were observed in surveys of sites in the adjacent Mystic and Poquonnock Rivers.A recreational boating and fishing pier in the lower Thames River (~16km distance from Stonington Harbor; site 15/16, Annex 1) was found to have low densities (~ 1% cover) of C. lepadiformis.Underwater photographs taken December 2008 at this pier also revealed the presence of the species, indicating that the population appeared to have over-wintered.During 2008, other sites in the lower Thames River had been surveyed, none of which had C. lepadiformis present (Annex 1).

Figure 3 .
Figure 3. Maximum likelihood tree of CO1 (369 bp).All branches with greater than 50 percent bootstrap support are labeled.Sequences of C. lepadiformis from the Atlantic clade (Turon et al. 2003) are indicated with a bracket.Scale bar indicates number of substitutions/site.

Table 1 .
Model parameters used in Maximum Likelihood analysis.