New records of the rare deep-water alga Sebdenia monnardiana (Rhodophyta) and the alien Dictyota cyanoloma (Phaeophyceae) and the unresolved case of deep-water kelp in the Ionian and Aegean Seas (Greece)

ROV and diving surveys

Macroalgal surveys were conducted from the R/V Philia (Hellenic Centre for Marine Research) from March 10–17 to October 20–25, 2013. Surveys utilised a remotely operated vehicle (ROV, SAAB Seaeye Falcon, rated to 300 m depth; www.seaeye.com) equipped with a video camera and robotic arm capable of collecting sessile organisms. In addition, scuba diving and snorkelling in adjacent infralittoral waters were performed, using the dingy of the R/V Philia as a support platform. Target areas for Laminaria rodriguezii had been chosen using maps (Figure 1) based upon the prediction model of deep-water kelp habitats based on water transparency and temperature data, previously used in tropical seas (Graham et al. 2007). The model combined satellite-derived data on water transparency and incident irradiance with locally-derived data on bathymetry and water temperature depth profiles to estimate the extent of substrate with enough light to support kelp productivity that existed below the mixed layer depth. This area would be considered the most likely to support deep L. rodriguezii populations, regardless of the quality of the substratum (i.e. soft- vs. hard-bottom).

Figure 1:

Prediction maps of the potential occurrence of Laminaria rodriguezii in the Eastern Mediterranean.

Coloured shading indicates potential deep-water kelp habitat colour-coded by depth range. These grid cells met all model criteria for kelp occurrence. Grid cells are approximately 3.8×3.8 km². Predicted kelp habitat shallower than 50 m is not shown. Black crosses show locations of predicted kelp habitat deeper than 110 m. Grey shading shows ETOPO2v2 bathymetry from 0 m (dark grey shading=land) to the 200 m isobath (lightest grey shading). The numbers correspond to the ROV surveys detailed in Table 1. O denotes the Jabuka area, where L. rodriguezii has been documented (Žuljevic et al. 2016), while asterisks denote the sites nearest to the ones reported in this study where Dictyota cyanoloma was found. (A) Central Mediterranean. (B) Eastern Mediterranean.

Surveys in March 2013 spanned the waters from Patras and Kefalonia in the south to Corfu (Kerkyra) in the north, while the surveys in October 2013 extended from Aidipsos (Euboia Island) towards Oreoi, along the mainland coast of Pelio and around the Northern Sporades Islands (Figure 1 and Table 1 for all stations of ROV surveys). These were complemented by snorkelling surveys at sites on adjacent coasts, in particular the surroundings of Argostoli and the coast of southeast Kefalonia, aiming to study the overall shallow algal biodiversity.

Collection of materials

Seaweed specimens were collected as entire thalli either by the robotic arm of the Seaeye Falcon™ ROV, by snorkelling or by SCUBA diving. They were subsequently conserved as herbarium specimens on Bristol paper or on microscope slides using acetocarmine as fixative and dye and 50% Karo™ syrup as mounting medium (Müller and Ramírez 1994).

Seaweed tissue samples were conserved in parallel in Silica gel and cetyltrimethylammonium bromide (CTAB; e.g. Gachon et al. 2009) buffer for further molecular study. Representative herbarium specimens and permanent slides were deposited at the East Mediterranean Seaweed Herbarium at the Hellenic Centre for Marine Research (HCMR) in Anavyssos, Athens.

DNA extraction, PCR amplification and sequencing

DNA extractions were carried out using the DNeasy Plant Mini Kit (Qiagen, Hilden, Germany) on material from specimens preserved in silica gel. From Dictyota cyanoloma, the partial mitochondrial cytochrome oxidase I (COI) was amplified using either the primer pair COI-789F & COI-1378R (Silberfeld et al. 2010) or cox1-GazF1 & cox1-GazR1 (Saunders 2005).

PCR amplification was performed in a total volume of 25 μl, containing 1.25 units μl⁻¹ (0.25 μl stock) of Taq DNA Polymerase (Promega), 5 μl GoTaq™ buffer, 5 mm MgCl₂, 1.25 mm dNTPs (desoxy nucleotides), 1.87 mm of each primer and 1 μl of template DNA (5–50 ng μl⁻¹). For Dictyota cyanoloma, PCR amplification was conducted with the primer pair cox1-GazF1/cox1-GazR1 (Saunders 2005). PCR amplification was carried out with an initial denaturation at 94°C for 2 min; followed by 35 cycles of amplification consisting of denaturation at 94°C for 30 s, annealing at 50°C for 30 s and elongation at 72°C for 1 min, followed by a final extension at 72°C for 5 min. Returned chromatograms (in ABI/Applied Biosystems format) were imported into BioEdit (Hall 1999) for quality control and resulting consensus sequences were queried against the GenBank online database using BLASTn (Altschul et al. 1997).

For Sebdenia monnardiana, the markers rbcL, 3′-COI and SSU were used. rbcL was amplified using the primer pair RbcL77F (Yang et al. 2014) – RbcL952R (Kawai et al. 2007), as Ral-R952), 3′-COI was amplified with primers cox1-789F and cox1-1378R (Silberfeld et al. 2010), and SSU with primers NS1F-NS4R (White et al. 1990), AFP2F (Burkhardt and Peters 1998) and AFP1R (5′-GGTAATGATCCTTCCGCAG-3′). PCR was performed in a total volume of 25 μl, with 21.5 μl of Taq Master Mix stock (VWR International, Haasrode, Belgium), 1.0 μl of each primer (10 mm) and 1.5 μl of template DNA. For primers RbcL77F-RbcL952R, PCR included an initial denaturation at 94°C for 3 min, followed by 35 cycles of denaturation at 94°C for 1 min, annealing at 48°C for 30 s and elongation at 72°C for 1 min, followed by a final extension at 72°C for 5 min. For primers NS1F-NS4R and AFP2F-AFP1R, PCR included an initial denaturation at 94°C for 3 min, followed by 35 cycles of denaturation at 94°C for 1 min, annealing at 50°C for 30 s and elongation at 72°C for 1 min, followed by a final extension at 72°C for 5 min. PCR success was confirmed by 1.2% agarose gel electrophoresis, and visualised by SYBR safe DNA gel stain (Invitrogen/ ThermoFisher Scientific, Waltham, MA, USA) under UV light. Acquired PCR products were purified using the GeneJET PCR purification Kit (ThermoFisher Scientific, Waltham, MA, USA) and sent off to be sequenced by an external Sanger sequencing service (Source Biosciences, Nottingham, UK). Returned chromatograms were curated similarly as with Dictyota cyanoloma.

Molecular phylogeny

Phylogenies for Sebdenia monnardiana were calculated separately for SSU and rbcL regions. Sequences including other relevant taxa of the Sebdeniales and Rhodymeniales (Table S1 for SSU, Table S2 for rbcL) were aligned and annotated in BioEdit (Hall 1999), and Maximum Likelihood (ML) trees were generated using MEGA7 (Kumar et al. 2016). The ML tree was generated based on the GTR model with Gamma distributed rates among sites, with invariant sites included. One hundred bootstrap replicates were used to assess tree robustness. Members of Rhodymeniales were used as outgroups in the analyses, Perbella minuta (Kylin) Filloramo et G.W. Saunders and Hymenocladia chondricola (Sonder) J.A. Lewis in the SSU phylogeny and Rhodymenia pseudopalmata (J.V. Lamouroux) P.C. Silva and Dictyothamnion saltatum A.J.K. Millar in the rbcL phylogeny.

Deep-water kelp in Greece?

Despite the survey efforts within the framework of the two cruises covered here, the question of the existence of Laminaria rodriguezii in east Mediterranean waters has to be considered unresolved. As of now, L. rodriguezii has never been observed in Greek waters. Previous records of Laminaria digitata (Hudson) Lamouroux and Saccharina latissima (Linnaeus) Lane, Mayes, Druehl & Saunders from the Ionian Sea cannot be verified due to loss of these materials during World War II. These possibly referred to Saccorhiza polyschides and not to deep-water kelps because they were based on specimens cast ashore together with seagrass leaves. The absence of observations during the cruises does not constitute conclusive negative evidence considering the patchy occurrence in areas such as the Adriatic where numeric prediction modelling would suggest a more contiguous occurrence (Žuljevic et al. 2016). Also, the intense bottom trawling activity in part of the Greek waters has to be taken into account. Even if L. rodriguezii had existed in any of the areas surveyed here, its communities may well have been exterminated in recent decades, in analogy to its disappearance from >95% of its previously-documented range in the Adriatic Sea due to bottom trawling (Žuljevic et al. 2016). Indeed, at several of the sites, the impacts of bottom trawling (trawling scars) were obvious during the ROV surveys.

Another hypothesis that could potentially account for the lack of Laminaria rodriguezii observations during the surveys reported here is that the model (Graham et al. 2007), which was developed and successfully tested for tropical locations (e.g. Galapagos), insufficiently reflects the strongly seasonal environmental conditions of the Mediterranean, and that the model may need further refinement. Specifically, although the model developed by Graham et al. (2007) predicted locations in our study region where deep L. rodriguezii may occur, if L. rodriguezii is less tolerant of lower irradiance levels than the physiological parameters used in the model, then the predicted areas in Greek waters may have been over-estimates of where L. rodriguezii could persist.

Without doubt, the ROV surveys conducted within the framework of this study have to be considered insufficient for conclusively answering the question of the occurrence of Laminaria rodriguezii in eastern Mediterranean waters. We suggest that fishermen, ROV operators and divers operating in the depth range of 40–150 m for unrelated purposes in the eastern Mediterranean should be given adequate instructions to be able to recognise this species in case of chance encounters.

Dictyota cyanoloma

The recently described Dictyota cyanoloma (Dictyotales, Phaeophyceae; Tronholm et al. 2010) is known from the Mediterranean coast of Spain (Tronholm et al. 2010, Bárbara et al. 2015), the Adriatic, Portugal, Madeira, the Azores and Canary Islands (Tronholm et al. 2010). Its phylogenetic affinities have been established by sequences of partial LSU rDNA, rbcL, psbA, cox1, cox3, and nad1 (Tronholm et al. 2010). Recent findings from Turkey (Taskin 2013) constitute the first records for the eastern Mediterranean so far.

The record of this taxon from Kefalonia Island reported here supports the notion that the eastern Mediterranean remains incompletely surveyed for its macroalgal diversity. DNA sequences unambiguously confirmed that the material belongs to Dictyota cyanoloma. In the eastern Mediterranean, D. cyanoloma has been previously reported from the Aegean coast of Turkey near Izmir (Taskin 2013), but not from Greek waters. This record suggests a contiguous, albeit rare distribution of this taxon in the shallow infralittoral from the Azores, Madeira and the Canary Islands, and throughout the Mediterranean, including the Adriatic to the Mediterranean coast of Turkey. Recent work (Steen et al. 2017) has shown that this taxon is not native to Europe, but an introduction which has been present in the Adriatic as early as 1935. The same study (Steen et al. 2017) showed that D. cyanoloma is also present in western and southern Australia.

Sebdenia monnardiana

Sebdenia monnardiana is known only from circalittoral habitats of the western Mediterranean – but not yet from the eastern Mediterranean. No SSU sequences, only rbcL sequences are available for S. monnardiana so far.

The finding presented here constitutes a new record for Greece and, indeed, the entire eastern Mediterranean. The Sebdeniaceae was erected by Kylin (1932) in order to accommodate Sebdenia monnardiana originally described from the North African coast of the western Mediterranean. Athanasiadis (2002; p. 120) reported it to be “the largest Mediterranean red alga, reaching 1 meter in diameter”. Compared to this, the specimen reported here (of only about 10 cm size) is very small. The SSU sequence reported here is the first sequence of this important marker for S. monnardiana, which is significant since S. monnardiana is the type species for the genus Sebdenia and the order Sebdeniales (Withall and Saunders 2006), and it is clearly difficult to collect. Our rbcL sequence matches two sequences reported previously for this taxon from Italy (especially AY294395.1 with 99.53% identity for 100% query cover, but also U21600. 1, with 98.94% identity). The specimen corresponding to AY294395 was collected at Lachea Island, near Catania, Italy, by G. Furnari and M. Cormaci in October 1994, and is surely a correctly identified S. monnardiana (Gavio et al. 2005). The lower similarity value for U21600 is due to missing data and possible errors in this older sequence; nevertheless it forms a clade with AY294395 and our sequence. Sequence similarity of our specimen with U21600 and AY294395 confirms the morphological identification of our material. In contrast, the sequence EU543493 for an alga from central Italy is problematical. It was submitted to NCBI by R. D’Archino, N. Abdelahad, and G. Procaccini and is only 89.4% identical with EU543493 and 90.6% with our sequence. However, it appears to be a good rbcL sequence (most substitutions are at third base positions) and likely belongs to a related species of as yet unknown identity, which in our rbcL tree (Figure 4A) forms a clade with Crassitegula sp. from Australia.