Searching for Mediterranean bath sponges (Demospongiae: Dictyoceratida: Spongiidae) in the Northeast Atlantic reveals a new species: an integrative taxonomic approach Author Grenier, Marie Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE), UMR CNRS 7263, IRD 237 - Aix Marseille Université - Avignon Université, Station Marine d’Endoume, Chemin de la Batterie des Lions, 13007, Marseille, France & Neuralia SAS, 410 Chemin Départemental 60, 13120, Gardanne, France Author Simmler, Charlotte Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE), UMR CNRS 7263, IRD 237 - Aix Marseille Université - Avignon Université, Station Marine d’Endoume, Chemin de la Batterie des Lions, 13007, Marseille, France Author Chevaldonné, Pierre Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE), UMR CNRS 7263, IRD 237 - Aix Marseille Université - Avignon Université, Station Marine d’Endoume, Chemin de la Batterie des Lions, 13007, Marseille, France Author Callizot, Noëlle Neuralia SAS, 410 Chemin Départemental 60, 13120, Gardanne, France Author Pérez, Thierry Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale (IMBE), UMR CNRS 7263, IRD 237 - Aix Marseille Université - Avignon Université, Station Marine d’Endoume, Chemin de la Batterie des Lions, 13007, Marseille, France thierry.perez@imbe.fr text Zoological Journal of the Linnean Society 2024 2023-11-14 202 1 23 journal article 306278 10.1093/zoolinnean/zlad166 bd3c20d5-db08-4d05-b5ab-d5bb9a339ae8 0024-4082 14340955 ED45D8A-954B-4723-9D0F-714357A7DAB5 Spongia lamella ( Schulze, 1879 ) ( Fig. 4 ) Material examined 20200602-MRS-MTP3, Pharillons de Maïre, cliff at 20 m depth, Marseille, France ( 43°12 ʹ 27″N , 5°20 ʹ 17″E ). Collector: T. Pérez, 2 June 2020 . 20200602-MRS-MTP4, Pharillons de Maïre, cliff at 20 m depth, Marseille, France ( 43°12 ʹ 27″N , 5°20 ʹ 17″E ). Collector: T. Pérez, 2 June 2020 . 20200602-MRS-MTP5, Pharillons de Maïre, cliff at 20 m depth, Marseille, France ( 43°12 ʹ 27″N , 5°20 ʹ 17″E ). Collector: T. Pérez, 2 June 2020 . 20200602-MRS-MTP6, Pharillons de Maïre, cliff at 20 m depth, Marseille, France ( 43°12 ʹ 27″N , 5°20 ʹ 17″E ). Collector: T. Pérez, 2 June 2020 . 20200602-MRS-MTP7, Pharillons de Maïre, cliff at 20 m depth, Marseille, France ( 43°12 ʹ 27″N , 5°20 ʹ 17″E ). Collector: T. Pérez, 2 June 2020 . 20200602-MRS-MTP8, Pharillons de Maïre, cliff at 20 m depth, Marseille, France ( 43°12 ʹ 27″N , 5°20 ʹ 17″E ). Collector: T. Pérez, 2 June 2020 . 20160607-GIJ-JC1, Las Gemelas island, rock substrate at 12 m depth, Gijon, Spain ( 43°33 ʹ 22″N , 5°36 ʹ 31″W ). Collector: J. Cristobo, 7 June 2016 . 20200624-HEN-BMNC2, Belhara, on reef flat at 22 m depth, Hendaye, France ( 43°24 ʹ 00″N , 1°43 ʹ 03″W ). Collector: M. N. de Casamajor, 24 June 2020 . 20200814-HEN-AMNC3, Abbadia, on reef flat at 13 m depth, Hendaye, France ( 43°23 ʹ 56″N , 1°45 ʹ 28″W ). Collector: M. N. de Casamajor, 14 August 2020 . 20210630-CIB-BPC1, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: P. Chevaldonné, 30 June 2021 . 20210630-CIB-BPC3, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: P. Chevaldonné, 30 June 2021 . 20210726-CIB-BTP1, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: T. Pérez, 26 July 2021 . 20210726-CIB-BTP2, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: T. Pérez, 26 July 2021 . Figure 4. In situ pictures of Spongia lamella in the Northeast Atlantic Ocean and the Mediterranean Sea, A, Baloo tunnel, Ciboure, France. B, C, Mouro Island, Santander, Spain. D, Marseille, France. 20210726-CIB-BTP3, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: T. Pérez, 26 July 2021 . 20210726-CIB-BTP4, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: T. Pérez, 26 July 2021 . 20210726-CIB-BTP7, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: T. Pérez, 26 July 2021 . 20210726-CIB-BTP10, Baloo tunnel at 14 m depth, Ciboure, France ( 43°24 ʹ 59.0″N , 1°39 ʹ 52.2″W ). Collector: T. Pérez, 26 July 2021 . 20210729-ESP-MTP5, western part of Mouro island, semi-dark to dark cave, between 3 and 12 m depth, Santander, Spain ( 43°28 ʹ 22.0″N , 3°45 ʹ 21.7″W ). Collector: T. Pérez, 29 July 2021 . 20210730-ESP-M2TP2, eastern part of Mouro island, shallow water cliffs and chaos of rocks and crevices, between 3 and 12 m depth, Santander, Spain ( 43°28 ʹ 21.1″N , 3°45 ʹ 18.3″W ). Collector: T. Pérez, 30 July 2021 . 20210730-ESP-M2TP3, eastern part of Mouro island, shallow water cliffs and chaos of rocks and crevices, between 3 and 12 m depth, Santander, Spain ( 43°28 ʹ 21.1″N , 3°45 ʹ 18.3″W ). Collector: T. Pérez, 30 July 2021 . 20210730-ESP-M2TP12, eastern part of Mouro island, shallow water cliffs and chaos of rocks and crevices, between 3 and 12 m depth, Santander, Spain ( 43°28 ʹ 21.1″N , 3°45 ʹ 18.3″W ). Collector: T. Pérez, 30 July 2021 . Comparative material examined Spongia virgultosa (Schmidt, 1868) 20070718-MRS-ECITP5, Jarre, cave in the semi-dark part at 15 m depth, Marseille, France ( 43°11 ʹ 46″N , 5°21 ʹ 55″E ). Collector: T. Pérez, 18 July 2007 . Coscinoderma sporadense Voultsiadou-Koukoura, van Soest & Koukouras, 1991 Po.25932, Tel-Aviv , Israel ( Idan et al. 2018 ). Description This sponge can be massive and irregular, with lobes or protuberances, fan- or vase-shaped when it becomes very big ( Fig. 4A–D ). The largest studied specimens measured 50 cm in their largest diameter, but this dimension can be> 1 m in the Mediterranean Sea ( Fig. 4D ). No true vase- or fan-shaped specimen was observed among our Atlantic specimens. This sponge presents a light to dark grey external colour, and the internal tissue is usually light or tawny yellow. The colour does not change upon fixation in 95% ethanol. The sponge consistency is flexible and hardly tearable. The surface is finely covered with small, evenly distributed conules. Oscules can be grouped in large individuals, or more randomly distributed in smaller ones, always on the outer sides in large individuals or on the top of each lobe or protuberance in smaller ones. They measure 1–3 mm in diameter and are slightly raised, with a small whitish membrane delimiting the opening. The ectosomal skeleton harbours an epidermal skin, 150– 600 µm thick, easily detachable, with a star-like appearance. It is made of a network of abundant foreign spicules and debris ( Fig. 5A–C ). Small canals can be observed under the surface. In the choanosome, debris and foreign spicules are still abundant at the periphery of canals, and thus throughout the sponge body ( Fig. 5D ). The primary fibres, 30–120 µm in diameter, are common, simple, irregular, and cored with numerous inclusions of foreign spicules and debris ( Fig. 5A, B ). Near the surface, they can be ramified. Secondary fibres, 20–40 µm in diameter, without inclusions or pith, form a dense ramified and irregular network ( Fig. 5B, E ). Tertiary fibres are present or absent. When they exist, their thickness is ≤10 µm in diameter. Distribution Mediterranean Sea and South European Atlantic Shelf (e.g. Noyer and Becerro 2012 ). This study expands the distribution of S. lamella northwards. Ecology In the Mediterranean Sea, S. lamella is well known to be distributed mainly along coralligenous cliffs or on deep mesophotic bottoms ( Vacelet 1959 , Pronzato and Manconi 2008 ). In the Northeast Atlantic, S. lamella has mostly been found in semi-dark and dark cavities, sometimes in crevices found along shallow water cliffs or under rocks. All specimens reported here were found in habitats with very high wave energy, which might explain why no large or fan-shaped individuals were observed. This sponge has a patchy distribution, and it lives in syntopy with S. maitasuna . In caves, this species is commonly associated with numerous other sponges, solitary scleractinians, such as Leptopsammia pruvoti Lacaze-Duthiers, 1897 ? , or zoanthids, such as Parazoanthus axinellae ( Schmidt, 1862 ) . No sign of epibiosis or predation was observed. No reproductive element was recorded in the processed specimens. Taxonomic remarks The 15 specimens examined from the Northeast Atlantic differ slightly in their growth form from S. lamella in the Mediterranean Sea, but their skeletons are identical. Openings of the oscula are also twice as large in the Atlantic specimens. They all present in their ectosome a rather thick layer of foreign spicules and debris, which is an unusual structure among Spongia that was not observed in S. maitasuna . Moreover, the overall network of fibres cannot be confounded with the specific honeycomb-shaped skeleton of the new species. A rather organized ectosomal skeleton made of foreign spicules and debris is known only in Coscinoderma and Leiosella , but in these cases, this structure is much thicker and the organization of the choanosome skeleton very different (e.g. intertwined secondary fibres in the Mediterranean C. sporadense ). Atlantic and Mediterranean specimens of S. lamella share the same common simple and irregular primary fibres, with inclusions of foreign spicules and debris, easily observable from the inner part of the sponge to the surface. They remain simple until the formation of conules, whereas in other Spongia species they become anastomosed at this level. The size of the primary fibres appears a little more variable in the Atlantic specimens (30–120 µm in diameter) than in the Mediterranean specimens (50–100 µm in diameter). Foreign spicules and debris seem to be more abundant in Atlantic specimens. In comparison, the primary fibres of S. virgultosa , which has an unusual encrusting papillate growth form for Spongia , are also full of mineral debris, but they are rare and significantly thinner (40–50 µm in diameter) than in S. lamella . Therefore, the slight morphological differences recorded between Atlantic and Mediterranean specimens (growth form, abundance of debris, and variation in the size of primary fibres) do not seem to us to be sufficient to consider the Atlantic specimens as a different species and are thus interpreted as variations attributable to the different environmental contexts. DNA analysis of studied species GenBank accession numbers for all sequences generated or downloaded in this study are presented in Table 1 . For the 28S (C2–D2 fragment) and CO1 (dgLCO1490–dgHCO2198 fragment) molecular markers, we obtained, respectively, 18 and 3 sequences of S. maitasuna , 21 and 14 sequences of S. lamella , 14 and 5 sequences for S. officinalis , 8 and 2 sequences of S. mollissima , 4 and 3 sequences of H. communis , 1 and 1 sequence of S. virgultosa , and only 1 sequence of S. zimocca (28S) and 1 sequence of C. sporadense (28S). Molecular markers allowed us to build highly congruent trees, although the phylogenetic reconstruction with 28S gave a better resolution ( Fig. 6 ) than CO1 (Supporting Information, Fig. S1 ). Both methods (NJ and ML) of phylogenetic reconstruction using 28S produced the same topology: 70 Spongiidae specimens divided in two, with only one sequence of C. sporadense on one side and 69 specimens of Spongia and Hippospongia on the other side ( Fig. 6 ). This set of specimens can be divided into three different groups. The first group is well supported by both methods of phylogenetic constructions (NJ and ML). It is composed by the proposed new species S. maitasuna , individuals from the two sampled localities presenting the same sequence of 28S, and by a well-supported group made of one sequence of S. zimocca , two downloaded sequences of S. nitens , and three sequences obtained in the present study of S. cf. officinalis from Ceuta . Spongia maitasuna thus appears closely related to this group, with the p-distance being very low between the proposed new species and S. zimocca (1.05%, four substitutions), S. nitens , and S. cf. officinalis (0.79%, the substitutions). The second group is weakly supported, whatever the method of construction. It includes S. virgultosa and a well-supported group of all S. lamella from the Mediterranean and Northeast Atlantic. With only one exception,thesequencedindividualsaregroupedbygeographical origin. The Mediterranean sequence (20200602-MRS-MTP8) mixed with the Atlantic presents a p-distance of 0.26% (one substitution) from the rest of the Mediterranean S. lamella . In comparison, S. virgultosa presents a p-distance of 13.4% (51 substitutions) from all S. lamella . The third group is composed by a weakly supported set of sponges, including all S. mollissima , S. officinalis , and H. communis . The genetic distances between these three species are low, with a minimal p-distance of 0.26% between S. mollissima and H. communis from Tunisia , and 0.79– 1.57% (three to six substitutions) between S. mollissima and S. officinalis or H. communis . Figure 5. Sections of the spongin skeleton of Spongia lamella . A, general view of the surface and the ectosome, showing the crust made of foreign spicules and debris at the surface, and the simple primary fibres. B, primary fibres including foreign spicules and debris, and secondary fibres forming an irregular and dense network. C, tangential section of the surface crust. D, foreign spicules and debris around canals. E, details of the secondary fibres. A phylogenetic tree was made by concatenating all sequences available for both 28S and CO1 ( Fig. 7 ), with the exception of two samples of S. zimocca and H. communis for which only 28S could be amplified. The topology of the concatenated tree is congruent with the 28S tree. Both phylogenetic reconstruction methods used here (BI and ML) presented the same pattern. Thirty specimens are grouped into two different groups representing the Spongiidae family. Group 1 is composed by one specimen of the genus Coscinoderma , whereas group 2 is composed by 29 specimens of Spongia and Hippospongia . The proposed new species, S. maitasuna , forms a well-supported group in both phylogenetic analyses and still separate but relatively close to S. zimocca , S. nitens , and S. cf. officinalis from Ceuta . All S. lamella form a well-supported group, which still appears separate by geographical origin. In this representation, S. officinalis and H. communis appear mixed in a subgroup also containing S. mollissima . Chemical fingerprints of the studied species We evaluated whether Spongiidae could be organized and grouped based on their chemical fingerprints and whether this chemical-based classification was congruent with the molecular phylogeny. Using the first aligned feature list, containing 915 chemical signals, we performed hierarchical clustering analysis to group sponge extracts according to their chemical similarities. The dendrogram ( Fig. 8 ) shows distinct groups of extracts all attributed to three distinct groups of species. The first group (group 1) includes all specimens of the proposed new species, S. maitasuna , in a subgroup, and S. nitens and S. zimocca in another subgroup. No sample of S. cf. officinalis from Ceuta (Strait of Gibraltar ) was available for this analysis. The second group (group 2) includes all specimens of S. lamella , still separated into two subgroups according to their geographical origin. The Mediterranean exception (20200602-MRS-MTP8) that was grouping in the 28S phylogenetic with Atlantic specimens is grouped here with all other Mediterranean samples of the same species. The third main group (group 3) includes S. officinalis , S. mollissima , and H. communis , but in the case of this metabolomic analysis, all species formed well-separated subgroups. Thus, although its typology cannot be compared with those of the phylogenetic trees, the groupings presented in the dendrogram resulting from the metabolomic analysis are well congruent with those permitted by the morphological and genetic analyses. The heatmap ( Fig. 9 ) complements these results by depicting the whole set of chemical features that participated in the sample classification. This representation illustrates clearly that extracts from S. maitasuna contain a distinct set of chemical features compared with other Spongiidae samples. Next, we assessed the distribution of features within the three identified groups of species to evaluate the chemical diversity and number of unique chemical signals characterizing each species. To that end, a second aligned feature list was generated from pooled extracts from each identified taxonomic unit and contained 240 features. Those features represent the most abundant and redundant chemical signals in each taxonomic unit. Venn diagrams were constructed to depict the number of chemical features shared between species as opposed to those that were detected in one or two species ( Fig. 10 ). The lowest number of features was detected for group 1, for which 45 chemical signals were shared between all three species ( S. maitasuna , S. nitens , and S. zimocca ). A total of 16 unique signals were detected for S. maitasuna , illustrating a much higher chemical diversity and uniqueness in extracts than in S. nitens (2) and S. zimocca (1). At this stage, none of these features has been identified formally, because the purpose of this analysis was not to assign the associated molecular structure to each feature. However, it is certain that none of the well-known and referenced compounds from groups 2 and 3 (as shown later) were detected among the signals. The highest number of features was associated with group 2, represented by S. lamella from different locations (Mediterranean and Atlantic). There were between 80% and 85% similarities of detected signals ( 111 in total) in all extracts. Among the detected chemical features, nitenin is one of the most abundant and well-known furanoterpenoids from this species ( Noyer et al. 2011 ). Extracts from the Mediterranean specimens present a greater number of unique detected signals (13) than those from the Atlantic specimens (5 and 4). Group 3, represented by S. officinalis , S. mollissima , and H. communis , also contains a relatively high number of detected features compared with the proposed new species. Extracts from these three species have 98 chemical signals in common. Some other well-known and characteristic furanoterpenoid compounds could be identified among these signals, such as furospongin-1, furospongin-4, and demethylfurospongin-4 ( Bauvais et al. 2017 ). Extracts from S. mollissima have the highest number of unique detected signals (23), compared with S. officinalis (3) and H. communis (9). Among all Spongiidae , S. lamella is the species presenting the highest chemical diversity, with a total of 170 chemical signals (including its unique chemical signals and others shared with other Spongiidae ). In comparison, extracts from the proposed new species contain 151 chemical features, and>10% of them are unique and thus have unknown molecular structures. After undergoing purification and determination of their chemical/molecular structure, they have the potential to become chemotaxonomic markers. Therefore, the hierarchical classification of species, based on their chemical fingerprints, is congruent with the phylogenetic trees previously presented. Each species can be differentiated by its chemical fingerprints associated with a unique set of chemical features that might serve as chemotaxonomic markers. These putative new phenotypic traits would need to be annotated or identified after isolation, purification, and structural analysis by means of additional spectroscopic (e.g. nuclear magnetic resonance) and spectrometric methods.