Morphological analyses were conducted under stereomicroscope Leica M205C with the use of Leica Application Suite v. 4.13. The following traits were measured: colony size, branch length, hydranth length, hydranth width, pedicel width, and number of tentacles.

Squash preparations for light microscopy were made using fresh nematocysts obtained from a living colony of C. cerulea. The nematocysts were examined and photographed using interference-contrast optics on a Nikon Eclipse Ti-S microscope. Morphometric measurements of hydranths were performed in vivo.

Genomic DNA was extracted from two colonies using DNeasy Blood and Tissue Kit (QIAGEN) following manufacturer’s instructions. Two mitochondrial markers, 16S rRNA and COI, and nuclear 18S rRNA were amplified using standard primers: SHA/SHB for 16S rRNA (Cunningham and Buss 1993), LCO-1490/HCO-Med-2414 for COI (Folmer et al. 1994; Ortman et al. 2010) and MitchA/MitchB for 18S rRNA (Medlin et al. 1988). The following PCR profiles were used: for 16S rRNA, 30 cycles of 20 s at 94 °C, 45 s at 50 °C, 120 s at 68 °C (Schuchert 2018); for COI, 40 cycles of 60 s at 94 °C, 120 s at 45 °C, 180 s at 72 °C (Ortman et al. 2010); for 18S rRNA, 30 cycles of 30 s at 94 °C, 60 s at 40 °C, 60 s at 72 °C, 300 s at 72 °C (Dunn et al. 2005). Amplified DNA was sent to Eurofins Genomics for Sanger sequencing. Consensus sequences were assembled in Geneious Prime (v. 2025.0.2) and deposited in GenBank (see Suppl. material 1).

Phylogenetic reconstruction was based on data previously reported in Mendoza-Becerril et al. (2018). We retrieved complete matrix (16S, COI and 18S) for all species reported in their study belonging to: Filifera, Filifera III (excluding Janaria mirabilis; renaming Lepidopora microstylus to Leptohelia microstylus, due to species status), Pandeidae (except Pandea sp.), Pseudothecata (renaming a likely mis-identified Podocorynoides minima as Cytaeis uchidae, Peter Schuchert personal communication), and additionally of Dicoryne conybearei (Suppl. material 1), totaling to 29 species. Sequences of each gene were then aligned in Geneious Prime with MAFFT using FFT-NS-i algorithm with 1000 iterations and 1PAM/k=2 scoring matrix and concatenated into a single matrix. Maximum likelihood phylogenetic analysis was run in W-IQ-TREE (Trifinopoulos et al. 2016) on concatenated multi-gene dataset with ModelFinder for model selection, FreeRate heterogeneity, edge-linked partition, and subsequent testing with ultrafast bootstrap analysis with 1000 iterations. Consensus tree was visualized in FigTree v.1.4.4 with additional formatting done in CorelDraw 2021.

Occurrence records of C. cerulea were downloaded from GBIF.org (2025) and OBIS (taxonid: 292221; accessed on 04-Feb-2025). Additional records were sourced from earlier taxonomic works (Suppl. material 2), most notably from species distribution reviews by: Vervoort (1964; European waters), Calder (2019; west Atlantic ) and Zaitsev and Öztürk (2001; Aegan, Marmara, Black, Azov and Caspian Seas). We also included references listed in the NEMESIS database (Fofonoff et al. 2018), excluding certain records due to misidentifications, inability to verify original sources or original location being not specific enough (Suppl. material 2). As literature-based records often lacked geographic coordinates, their position was approximated based on the descriptions given in the text (Suppl. material 2). All occurrence records were processed in R with tidyverse (v.2.0.0), to combine them into a single database and to remove duplicates. Then, maps were plotted in ArcGIS Pro (v.3.4.1).

To facilitate subsequent phylogeographic analyses, as well as to provide reference data for biological invasion monitoring, we present the first set of molecular data for C. cerulea that was used to reconstruct its phylogenetic position (Fig. 3, Suppl. material 1). The recovered topology of the majority of athecate hydrozoans (order Anthoathecata) was consistent with recent molecular works (e.g., Prudkovsky et al. 2017; Mendoza-Becerril et al. 2018). We documented paraphyletic relationships within Cytaeididae, Pandeidae, Oceanidae, indicating a non-monophyletic Bougainvilliidae with multiple divergent lineages scattered across the phylogeny (Fig. 3). We also found that sequences of C. cerulea clustered together with B. vestita and C. caspia. Although the branch support for this grouping was relatively low (79), this result suggests that they all should be united under the same family Cordylophoridae. This, however, will require a more thorough examination of their morphological and molecular diversity.

The original description of C. cerulea dates back to 1882 and was based on material from the western Atlantic, off the coast of Virginia, United States (Clarke 1882). In 1902, a similar species, Garveia franciscana, was described from the eastern Pacific, in San Francisco Bay, USA (Torrey 1902, as Bimeria franciscana). The distinction between these two species was based on the number of eggs produced by a gonophore, with a single egg developing in G. franciscana, and multiple eggs in C. cerulea. It was only recently, that Calder (2019), observed different developmental stages of gonophores, containing both a single egg cell, as well as multiple egg cells within the same colony. This led him to consider both species as conspecific, and to propose that they represent a single species under the priority name Calyptospadix cerulea (Calder 2019).

Records of C. cerulea span coastal regions in both hemispheres, extending as far north as the American subarctic (59°N) and as far south as the Australian coastline (27°S, Fig. 4A). This species has been widely reported from the Atlantic Ocean, including the USA east coast (Fraser 1944, as B. tunicata), the Gulf Coast of Florida (Calder 2019), and Venezuela (de Rincón and Morris 2003, as G. franciscana). It has also been recorded in European continental seas, such as the Mediterranean Sea (first record from Venice in 1978; Morri 1981, 1982), the Sea of Azov (Simkina 1963), the Black Sea (Paspalev 1933), the Caspian Sea (Zevina 1962), and the North Sea, especially it was very common in Zuiderzee, the no longer existing Dutch bay of the North Sea (Vervoort 1964, and references therein). Additional records come from more southern regions, including Australia (Pennycuik 1959; Davie 1990) and West Africa (Billard 1927; Buchanan 1956). Data from OBIS (2025) and GBIF (2025) also include tropical occurrences in the Indian Ocean and Caribbean Sea (Fig. 4).

In the Baltic Sea and adjacent waters, C. cerulea has been recorded only in the westernmost part of the basin (Fig. 4B), where it was previously identified under the name G. franciscana, including records from Kiel Bay (Thiel 1970), the Bay of Mecklenburg, and the Warnow Estuary (Penzlin 1957). We also suggest that Bimeria baltica, described by Stechow (1927) from the Greifswald Lagoon, may in fact be conspecific with C. cerulea. Although this identification was challenged by Kinne (1956b) and Penzlin (1957), Stechow’s description of the hydranth morphology, the arrangement of tentacles in a single apical whorl, their number, and the position of the gonophores closely match the diagnostic features of C. cerulea. Since then, this species has been reported multiple times along the German Baltic coast, including Kiel Bay (2016, 2020, as G. franciscana), Sassnitz on the island of Rügen, and the Bay of Mecklenburg (AquaNIS 2025). Further molecular work, based on specimens from the two type localities, as well as populations from other regions of the world are needed to better resolve the complex status of species considered synonymous to C. cerulea.

The low number of C. cerulea records since it was first reported by Stechow in 1927 may be due to the limited availability of natural hard-bottom habitats in the southern Baltic Sea (Tęgowski 2005). However, future occurrences may increase because of the growing number of artificial marine structures, e.g., offshore wind farms, cables, and underwater infrastructure (Kubacka et al. 2024), which can serve as “stepping-stones” facilitating the spread of non-native species. Additionally, intensified maritime traffic and inadequate ballast water management may further contribute to the introduction and expansion of C. cerulea in the region.

The expansion of C. cerulea in the Baltic Sea could have both ecological and economic consequences. Its detrimental impact on industrial water systems is well-documented in several locations worldwide. For example, dense colonies of this species have been found overgrowing the intake structures of cooling water systems at the Chesapeake Bay Nuclear Power Plant, clogging filtration screens and leading to a noticeable decline in pump efficiency (Tamburri 2014). In the Sea of Azov, C. cerulea has been identified as a major component of biofouling communities on marine industrial water intake structures at the Azovstal metallurgical plant in Mariupol, Ukraine (Simkina 1965). The abundant growth observed on water conduits suggests that rapid water flow creates particularly favorable conditions for its development (Simkina 1963). The expanding presence of C. cerulea has become a growing concern for industries in Venezuela that rely on Lake Maracaibo’s water. As a primary contributor to biofouling and biological corrosion, this species creates encrustations and causes blockages in submerged equipment and infrastructure (de Rincón and Morris 2003), resulting in substantial maintenance costs and increased efforts to prevent fouling and corrosion. In Chesapeake Bay, it presents a significant challenge for crabbers, as it heavily encrusts their traps, reducing their effectiveness. As a result, fishermen are forced to frequently clean and dry their gear (Andrews 1973). Similarly, it has become problematic for oyster farms, where it extensively overgrows both oyster trays and the oysters themselves (Andrews 1973).

Nevertheless, the positive role of three-dimensional, erect, bushy colonies of C. cerulea as habitat providers should not be overlooked. This species frequently serves as secondary substrate for various epibionts and may be densely colonized by protozoans. Notably, despite sampling taking place in winter, juvenile bivalves and Aurelia aurita polyps were also observed inhabiting its colonies, indicating that C. cerulea can offer refuge and settlement surfaces even during less biologically active seasons. Moreover, C. cerulea was observed to co-occur with the native G. loveni without signs of competitive exclusion, suggesting a degree of ecological compatibility between the two hydrozoan species.

Similar habitat functions of bushy hydroid colonies have been documented in other studies. Numerous amphipods, mud crabs, and various microscopic protozoans were found inhabiting dense colonies on experimental panels in the Patuxent Estuary (Cory 1967). Likewise, Simkina (1965) reported mobile invertebrates such as copepods, isopods, amphipods, and polychaetes thriving among elevated hydroid colonies. In fouling communities, hydroids also supported intense predatory interactions. For example, up to 7,000 Stiliger bellulus (A. d’Orbigny, 1837) nudibranchs per 100 g of hydroids were recorded feeding on colonies, while approximately 1,500 juvenile Rhithropanopeus harrisii (Gould, 1841) crabs were found hiding and feeding within a single square meter of hydroid-covered substrate (Simkina 1965).

Although more data is needed, the particularly broad tolerance of C. cerulea to both salinity and temperature make it very likely that this species will continue to thrive in the Baltic Sea. Given its potential to interact with native fauna and the still insufficiently understood ecological roles it may play, further research on C. cerulea is needed. A better understanding of its biology and ecological impact is essential to anticipate possible long-term consequences on native assemblages and to inform effective monitoring and management strategies within the HELCOM-regulated Baltic Sea ecosystem.