A new clionaid sponge infests live corals on the west coast of India (Porifera, Demospongiae, Clionaida)

Coral reef ecosystems depend on the balanced interplay of constructive and destructive processes and are increasingly threatened by environmental change. In this context bioeroding sponges play a significant role in carbonate cycling and sediment production. They occasionally aggravate erosional processes on disturbed reefs. Like other coral ecosystems, Indian reefs have suffered from local and global effects. However, the systematic affiliation and diversity of many Indian bioeroding sponges and their infestation rates are largely confused or unknown. The present study describes a new bioeroding sponge species, Cliona thomasi sp. nov. from the central west coast of India. It belongs to the Cliona viridis species complex, displaying the key characters of tylostyles and spirasters, as well as harbouring photosymbiotic dinoflagellates. Specific morphological characteristics and molecular data from nrITS1 DNA and 28S rDNA distinguished C. thomasi sp. nov. from other known C. viridis complex and a number of Spheciospongia species. The historic sample of ‘Suberites coronarius’ from Mergui Archipelago (sensu Carter, 1887), but not from the Caribbean (sensu Carter, 1882), is conspecific with C. thomasi sp. nov. Cliona thomasi sp. nov. is locally very abundant, appears to be a key bioeroder, and thus regular monitoring of its abundance, distribution and infestation patterns is recommended.


Introduction
Tropical coral reefs are marine habitats well recognized for their rich biodiversity and ecosystem services (e.g. Moberg & Folke, 1999). The health, resilience, and persistence of these ecosystems rely on an intricate interplay of physical, chemical and biological components that govern constructional and erosional processes (e.g. Hoegh-Guldberg et al., 2007;Worm et al., 2006).
Bioerosion is one of these processes and plays a large role in carbonate-dominated habitats such as coral and mollusk reefs (Perry & Harborne, 2016;Schönberg et al., 2017a). Bioerosion is the degradation of hard materials by living organisms (Neumann, 1966), contributing the negative part of the biogenic cycling of calcium carbonate (CaCO3) on tropical coral reefs (Schönberg et al., 2017a). Globally, bioeroding sponges are the dominant macroborers (Schönberg et al., 2017b), i.e. macrobiotic endoliths that live in and erode CaCO3 substrates.
The taxonomy of bioeroding sponges is challenging. This is due to a high degree of morphological, and phenotypic variability (Calcinai et al., 1999;Hill, 1999;Hill & Hill, 2002;Schönberg, 2008), to the occasional lack of specific, unambiguous spiculation for some species groups, and in some cases to lost or inaccessible type material (Schönberg & Beuck, 2007;Schönberg, 2013). In consequence, some species groups of bioeroding sponges are taxonomically insufficiently resolved, and especially in the Indian Ocean. This includes the Cliona celata Grant, 1826 species complex, and the genera Cliothosa, Pione and Spheciospongia (Schönberg et al., 2017b). Indian Cliona aff. viridis, Cliona aff. celata, Pione and Cliothosa species were historically simply identified as species that were already described from the Atlantic, assuming a wide distribution or cosmopolitanism. However, more recent results suggest that cosmopolitanism is not a common trait in bioeroding sponges, and that "species" with distributions across different oceans are often complexes of cryptic species with similar morphology (Boury-Esnault et al., 1999;de Paula et al., 2012;Xavier et al., 2010). A number of Spheciospongia species were described specifically from India (often as Spirastrella species; e.g. Dendy, 1905Dendy, , 1916, but this genus is also badly resolved, because the genus is quite diverse in the Indian Ocean, and its morphological characters can be highly variable (Schönberg et al., 2017b). Tylostyle dimensions in Spheciospongia species have a wide range of lengths due to dermal tylostyles often being only about half as long as choanosomal tylostyles (e.g. Dendy, 1905). Relying predominantly on spicule traits may thus be insufficient to characterise the local community of bioeroding sponges, and histological or molecular analyses are necessary (Bickford et al., 2006).
The occurrence of bioeroding sponge species in the Indian region is hence not well understood and requires a systematic analysis. In such cases, molecular taxonomy in addition to morphology may assist in identifying difficult or separating cryptic species (Bickford et al., 2006;Escobar et al., 2012;Leal et al., 2016;Xavier et al., 2010).
This publication aims to distangle some erroneous accounts of C. viridis complex species from western Indian coral reefs near Goa. The present systematic study includes both morphological as well as molecular analyses, and describes the new species C. thomasi sp. nov. that is closely related to Cliona orientalis.

Study sites
Specimens of bioeroding sponges were collected during 2015-2017 from the coral reefs of the Malvan Marine Sanctuary (site I; 15°58′N 73°30′E) and the Grande Islands (site II; 15°21'N and 73°45'E, Fig. 1) in the central eastern Arabian Sea. The sites represent two nearshore coral reefs at the central west coast of India, at a distance of approximately 54 nautical miles from each other. The sites also represent little-studied coral reef ecosystems in India.
The Malvan Marine Sanctuary is one of the seven marine sanctuaries in India and comprises a nearshore discontinuous reef, surrounding the north-eastern side of the Island and running parallel to the shore located 50 m to 200 m away from the shore (De et al., 2015) Numerous submerged, exposed rocks and the Sindhudurg Island provide suitable substratum and protected habitat for coral Formatted: Space After: 6 pt, Line spacing: 1.5 lines settlement and growth (Untawale & Dhargalkar, 2002). The site Grande Islands consists of two elongated islands, Ilha de São Jorge and Grande Island, which are located 2 km away from the coast (Mote, pers. obs.). Generic diversity of corals at Grande Island is relatively low in comparison to the other reefs of India (Manikandan et al., 2016).

Sponge collection and morphological studies
Bioeroding sponges were collected while SCUBA diving between 4 and 10 m water depth from three different stations from each of the two study sites (Figs. 1-2; Supplement 1). Sponge fragments were removed using hammer and chisel, placed separately into pre-labeled and sealable plastic bags and then immediately preserved in 96% ethanol at the field site. During collection, careful observations were made about the phenotypic characters of the sponge samples, such as macromorphology, color and substrate. The preserved samples were brought to the laboratory on ice and studied at the CSIR-National Institute of Oceanography, Goa. In the laboratory, sponge-infested substrata were broken open and studied under a stereomicroscope (SZX10, Olympus, Gurgaon, India). The detailed morphological analyses inclusive of the skeleton arrangement, spicule types and dimensions were made following Rützler (1974) and Schönberg (1999). Briefly, spicule preparations for both light and scanning electron microscopy (SEM) were obtained after 12h digestion of sponge tissue with 70% nitric acid heated to 80 º C. Spicules were rinsed with distilled water and dehydrated in 96% ethanol, concentrating the spicules partly by sedimentation, partly by centrifugation before carefully removing the supernatant. The final spicule-ethanol suspension was carefully mixed, spread, dried and mounted on microscope slides (in DPX mountant, Loba Chemie, Mumbai, India; or Eukitt, Sigma Aldrich, Sydney, Australia). Spicules were similarly spread and dried on SEM stubs, but without adhesive, and gold-sputtered for 15-20 minutes in a compact vacuum coating system (12157EQ, SPI-Module Sputter Coater, Global Nanotech, Mumbai, India, or DSR1 desk sputter coater PM View, NewSpec, Myrtle Bank, Ausrtalia ). Stubs were viewed with a JMS-5800LV scanning electron microscope (Jeol, Peabody, USA), and SEM photographs were taken using a TM3030 Plus Tabletop Microscope (Hitachi, Singapore). Viewing with light microscopy and micrometer eyepiece generated spicule dimensions (Olympus bright field microscope, Gurgaon, India; DMLED microscope, Leica, Macquarie Park, Australia). In four specimens 100 tylostyles were measured for biometric analyses (for our specimens MGB 21, MGB 23, MGB 33 and MGB 35, the latter two being part of the type series; Supplements S2-S3). Depending on specimen and comparisons needed, dimensions for taxonomic comparisons were then obtained as tylostyle maximum length and shaft width, as well as tyle width and length, and spiraster total length including spines, width without spines, and a count of bends. These were measured for 20 spicules of each spicule type per sponge specimen, if enough spicules could be found on the slide. We aimed to include only fully formed megascleres and to ignore slim and unfinished tylostyles (following methods in Schönberg & Beuck, 2007). For our target species from West India this meant to exclude most spicules of <12 µm shaft width. In comparatively short tylostyles the shaft width was proportionally slimmer than in long tylostyles, and overall we accepted tylostyles with a length : width ratio of <30 as fully formed spicules, even if they were occasionally slimmer than 12 µm. In this way we measured the first 20 tylostyles that were not obscured or broken. Relying on fully formed tylostyles generated quite stable and uniform means among different specimens of the same species and reduced the overlap of spicule dimensions between different species (Supplements S4-5) provide raw data and descriptive statistics in summary).
We compared spicule characters of the Indian material with representative collection vouchers (Supplement S6). This comparative material included subsamples of dried or wet specimens and slide preparations of sections, tissue plucks and spicules. Vouchers were from the second author's reference collection (CS), the British Museum of Natural History (BMNH), the Liverpool World Museum (LIVCM), and the Paris Natural History Museum (MNHN). After our analyses were completed, representative vouchers were selected from the Indian sponges and kept as reference material. A type series was deposited and registered at the National Institute of Oceanography Taxonomic Reference Centre, Goa, India (NIO). Further material was housed in the reference specimen collection of the Biological Oceanography Division of the NIO, and is also accessible for loan requests.

DNA extraction, amplification, and sequencing
DNA was extracted from representative subsamples of the Indian sponge with a DNeasy kit according to the manufacturer's protocol (Qiagen, Hilden, Germany). DNA was amplified using the following primers: nrITS1 DNA 5'AAAAGTCGTAACAAGGTTTCCG3' and 5'TTGCGTTCAAAGACTCGATG3' for a forward and reverse copy, respectively (after Escobar et al., 2012); the D2 region of 28S rDNA with 5'AAGGTGAAAAGTACTTTGAAAAGA3' and 5'TCCGTGTTTCAAGACGGGTC3' for a forward and reverse copy, respectively (after Barucca et al., 2007 sequencer, Applied Biosystems, Foster City, USA), and the obtained chromatogram was edited using Formatted: Space After: 6 pt, Line spacing: 1.5 lines the ABI sequence scanner software 1.0v. All the sequences were deposited in GenBank (NCBI; Benson et al., 2013) under the accession numbers MG367332-MG367341 (note that our specimen field references start with "MGB" and have different numbers in comparison to the NCBI numbers that coincidentally start with "MG"; the numbers are matched in Table 1).

Alignment and phylogenetic analysis
All sequences were aligned with Clustal W with the default parameters (gap open = 15; gap extension = 6.66; gap distance = 4 (Thompson et al., 1994). The alignment of the sequences was checked for the potential occurrence of nuclear pseudogenes using the genetic code for invertebrate mitochondria and for no frame-shift mutations, which would indicate that these sequences originated from a non-functional gene region. Obtained sequences were compared to published data of clionaid sponges (Table 1), and we jointly analyzed our sequences with the NCBI sequences from other Cliona species in MEGA v6.06 (Tamura et al., 2013). The most suitable models of the molecular evolution for each gene and the concatenated data were determined using the best fit substitution model Tamura-Nei modelT as indicated by model test 2 based on the Akaike Information Criterion (AIC; Dariba et al., 2012). To assess whether some of the Indian samples formed a distinct monophyletic clade relative to congeneric samples from other locations, we reconstructed a phylogenetic tree in MEGA v6.06 using Maximum Likelihood (ML) with complete deletion of character positions containing gaps and missing data (Tamura et al., 2013). Evolutionary distances were computed employing the Kimura 2-parameter method (K2P; Kimura, 1980), and support for individual nodes was based on 1000 non-parametric bootstrap estimates (Felsenstein, 1985). The K2P distances were also used to compare levels of genetic differentiation between the sequences generated in this study and the published Cliona sequences.

General Observations
Within the material sampled for the larger project, ten specimens displayed morphology similar to C.
viridis complex species. All of them appeared to belong to the same species and were studied in more detail (Supplement S4). Other species dissimilar to C. viridis were observed, and a few specimens were presently included into the phylogenetic analyses (Figs. 3-4), but these species will be treated and published elsewhere. Within this extra material, sample MGB 20 was also close to the C. viridis complex, but did not cluster together with our core material, MGB 12 associated with Cliona rhodensis and had tylostyles consistent with Cliona celata complex species, while MGB 24 separated out distantly from all other Cliona species and obviously did not belong to the Clionaida,

Taxonomic description of Cliona thomasi sp.nov.
External morphology: Both, in α-and β-morphology in the field, with tendency to -morphology in thick specimens, but without fistular processes typical for  specimens. Papillate α sponges in Turbinaria and Favites spp., individuals up to 20-40 cm in total diameter. Papillae circular or oval, very small, 0.3-0.8 mm in diameter ( Fig. 2; Supplement S1). Encrusting  to  sponges forming patches of 60-100 cm in diameter, with epilithic tissue 0.5-3 cm thick (Fig. 2.1-2.2). Surface smooth. Texture hard and incompressible due to underlying coral skeleton. Live colour beige-brown to dark brown, in alcohol initially pale brown with green surface, later fading. Oscules in live  sponges lighter in colour than remaining surface, being pale yellow, occasionally with blue hues (Supplement S1). Choanosome pale yellow in all observed specimens.

Remarks
The tylostyle shape of Cliona thomasi sp. nov. and the lack of a clear size difference between dermal and choanosomal tylostyles suggested that our West Indian sponges in encrusting to massive growth form belonged to the genus Cliona, rather than to Spheciospongia ( Fig. 6; Supplements S4-5). We were thus able to immediately reject all Spheciospongia species reported from the Indian Ocean or West Pacific that we had included for comparison (Supplements S5-6). Consistently, these had longer choanosomal tylostyles than commonly occurring in Cliona species and a second, dermal tylostyle type that was about 60% of the length of the former. C. thomasi sp. nov. also appeared to have a shorter tylostyle type, but this differed not as strongly from the main tylostyle as in Spheciospongia spp. (ca. 80% of the length of the main tylostyles; Supplements S2-3) and was apparently restricted to the choanosome and not associated with the dermal palisade. Spheciospongia megascleres were mostly variable subtylostyles with subtle, elongated tyles very unlike the pronounced, round tyles observed in C. thomasi sp. nov.
Further, our samples clearly belong to the C. viridis (Schmidt, 1862) species complex. This is indicated by our molecular results, but also by a series of morphological and ecological characters such as bioeroding activity, the brown colour, the presence of tylostyles and slender spirasters with multi-split spines, and the occurrence of dinoflagellate photosymbionts (Rosell & Uriz, 2002;Schönberg, 2002b;Schönberg & Loh, 2005;. After coming to this decision, our specimens were then difficult to further identify to species level. On one hand similarities to other Indo-Pacific C. viridis species prevented conclusive decisions, on the other hand decisions presented in earlier publications added to the confusion or provided misleading information that required careful consideration. We quickly excluded the Indo-Pacific C. viridis-like species Cliona caesia Cliona minuscula Schönberg et al., 2006, because they are known only in -morphology and have no spirasters Schönberg et al., 2006;Supplement S6).
Moreover, C. caesia has light blue oscular collars, and C. minuscula has much smaller tylostyles than C. thomasi sp. nov. Cliona subulata Sollas, 1878 is another -morphology sponge. The original sample site of C. subulata is unknown, but the species is assumed to be from the Indo-Pacific (Schönberg et al., 2017b). However, spirasters in C. subulata are helical and have characteristic, long and discrete spines unlike those on the spirasters of C. thomasi sp. nov. (Sollas, 1878;Supplement S6). Cliona albimarginata Calcinai et al., 2005 has a very similar habit compared to C. thomasi sp.
nov., but its fully grown tylostyles are significantly slimmer, and the tyles are oval and elongated in C. albimarginata, but predominantly spherical in C. thomasi sp. nov. (Calcinai et al., 2005; Supplement S6). C. albimarginata spirasters display more irregular shapes and are less frequently Cshaped than in C. thomasi sp. nov. C. orientalis Thiele, 1900 also has a very similar habit in morphology, as well as very similar spicule dimensions and spicule shapes compared to C. thomasi sp. nov. (Thiele, 1900;Supplement S6). C. orientalis microscleres are predominantly helical spirasters, however, and only occasionally C-shaped. Considering that spicule dimensions can vary to some degree with environmental conditions (Bavestrello et al., 1993a(Bavestrello et al., , 1993bCárdenas & Rapp 2013;Mercurio et al., 2000;Valisano et al., 2012), initially we could not safely exclude C. orientalis.
Even though C. varians occurs in the Caribbean, it shares morphologic characters with C. thomasi sp. nov., especially when considering the microscleres. C. varians is known for its C-shaped spirasters or "anthosigmas", which as in C. thomasi sp. nov. is by far the most predominant form of the microscleres (Supplement S6;Figs. 6,). However, the tylostyles in C. varians are longer and slimmer than in C. thomasi sp. nov., with oval, elongated tyles that are often subterminal ( Fig.   7.4, 7.7; Supplements S5-6). These characters appeared to provide reasonable grounds to distinguish the two species despite the similarities of the microscleres, but as previous publications from India had reported C. varians, we carefully searched the historical literature for further information.
We became aware of a possible conspecifity with Carter's (1882, 1887) Suberites coronarius (Carter, 1882), which Dendy (1916) transferred to Cliona after he correctly recognised that the species was actively bioeroding (Supplement S6). The problem was that Carter's material did not refer to one, but to two species (see Annandale, 1915;Dendy, 1916;Thiele, 1900). His 1882 Caribbean material was synonymised with C. varians, a species name which had seniority . This rendered the name "coronarius" unavailable, even though Carter's (1887) specimen from Mergui Archipelago was clearly different from C. varians and still needed a name.
Earlier workers had recognised the similarities with C. orientalis and declared Carter's Mergui specimen and a very similar Dendy (1916) specimen from northwestern India as conspecific with this species (Annandale, 1915;Dendy, 1916;Thiele, 1900). This decision was followed by Thomas (1972Thomas ( , 1979Thomas ( , 1986, who worked again on similar sponges from the Palk Strait (Supplement S6). To further complicate things, Topsent (1918)  Spicular comparisons helped us to confirm that the Caribbean C. varians was different from our Indian material and that respective earlier reports for C. varians from the Indian Ocean should be considered as wrong (e.g. Calcinai et al., 2000;Immanuel et al., 2015;Namboothri & Fernando, 2012; Supplement S6). The tylostyles in C. varians are overall slimmer than in C. thomasi sp. nov.
( Fig. 7.4, 7.7), and the characteristic anthosigmas in the former species are more regular and uniform in overall shape, unlike in our samples, where the spirasters can commonly be J-or hook-shaped ( Fig. 6, 7.5-7.6). However, Carter's (1887) "Suberites coronarius" from Mergui was a very good match. An additional sample was located in the National Institute of Oceanography, Goa (Devi et al., 2011), and we also regard a report by Dendy (1916) from the Gulf of Katchchh as consepcific with C. thomasi sp. nov. Other reports for C. viridis complex species could not as easily be aligned with our material (Supplements S5-6), and molecular analyses became unavoidable in order to confirm or reject conspecificity with existing C. viridis complex species (Table 1).
We successfully extracted DNA from the specimens listed in Table 1. Regrettably, the historical museum material did not presently yield useful molecular data that would have allowed direct comparison with our material. Moreover, the nrITS1 phylogram led to ambiguous results.
Pione clustered together with C. viridis species, and C. delitrix Pang, 1973 and "C. laticavicola" (Pang, 1973) did not form a tight clade (Fig. 3), even though recent results showed that the latter two are conspecific (Chaves-Fonnegra et al., 2017). Results such as these can possibly be explained with sample contamination by the DNA of other species that were sampled during the same program.
Bioeroding sponge tissue is difficult to extract, and some species can mingle or occur in close vicinity of each other (Schönberg, pers. obs.). When plucking tissue out of the substrate it is difficult to know whether a disparate species neighbours the target material or whether the plucked material contains exclusively tissue from a single species. In addition, specific sequences in the ITS region often evolve faster than those of other biomarkers, so that resulting intragenomic variability can result in alignment problems (Vollmer & Palumbi, 2004;, and this marker may not be the best to elucidate taxonomic and systematic relationships within the Clionaida. The more conservative 28S region previously generated good results for clionaid sponges in this context (Barucca et al., 2007;Leal et al., 2015;Xavier et al., 2010), and we will more strongly rely on the respective phylogram (Fig. 4).
The molecular analyses based on 28S again confirmed that our West Indian sponges belonged to the C. viridis species complex, but in the same time their sequences differed from all comparative material available on GenBank and clustered separately (Figs. 3-4). This excluded not only the morphologically very similar, Indo-Pacific species C. orientalis, but also common Caribbean sponges such as C. caribbaea Carter, 1882, C. tenuis Zea andWeil, 2003 andC. varians, as well as the Atlanto-Mediterranean species C. viridis.
Taking molecular and morphological results into account together, we concluded that mainly due to the spicule dimensions and shapes, our material was most likely a new species that we here described. The most distinctive characters of C. thomasi sp. nov. include its very robust and straight tylostyles with round tyles and the predominantly C-shaped, very slim spirasters. Based on spicule observations we also regard Thomas' (1972Thomas' ( , 1979Thomas' ( , 1986) "orientalis" specimens from Palk Strait as conspecific. Dendy's (1916) "Cliona coronaria" from Okha in Gujarat, NW India appears to be conspecific as well, which was not further assessed, however. Reports of "Cliona varians" in India were also tentatively included as synonyms for C. thomasi sp. nov., assuming that respective identifications were likely based on the high frequency of C-shaped spirasters (Kiruna-Sankar et al., 2016;Raghunathan, 2015aRaghunathan, , 2015b; but these were mere reports and did not provide descriptions).
According to these findings, we can presently report a distribution of C. thomasi sp. nov. for Northwest to South India and the Mergui Archipelago, with possible additional occurrence around the Andaman and Nicobar Islands. Because of the convoluted circumstances of this taxonomicsystematic investigation, we summarised respective information in Supplement S6.

Discussion
It is now widely accepted that the guild of bioeroding sponges contains a number of insufficiently resolved species complexes with similar morphological characters (Schönberg et al., 2017b). As a consequence a number of species have traditionally, but erroneously been grouped under one name and were regarded as cosmopolitan or as having a wide distribution across different oceans (e.g. Xavier et al., 2010). With molecular taxonomy being increasingly used in addition to morphological studies, new species have been recognised and described, as well as morphological features identified that characterise and distinguish them within these difficult groups (e.g. Boury-Esnault et al., 1999). the Mediterranean C. viridis/nigricans were accepted as C. labiata, C. viridis and C. parenzani, but some workers still recognise older synonyms as possibly valid (Longo et al., 2017); and respective Indo-Pacific species are presently recognised as C. albimarginata, C. caesia, C. minuscula, C.
orientalis, and likely C. subulata and C. vallartense. All these species harbour symbiotic dinoflagellates that are thought to provide essential nutrients to their hosts (Fang et al., 2014;Weisz et al., 2010). This may in part explain the diversity of this group, the large average specimen size and fast growth rates, their competitive strength and their success in general (Schönberg et al., 2017b). C.
viridis complex species are as a rule among the most dominant and destructive macroborers on coral reefs (Schönberg, 2001;Schönberg et al., 2017b), and C. thomasi sp. nov. is abundant and aggressive as well.
We therefore think that like some other C. viridis complex species, C. thomasi sp. nov. can aggravate coral bioerosion where it is common. Should abundances of C. thomasi sp. nov. increase, this could cause a gradual phase shift from constructional to erosional conditions on local reefs.
Increasing abundances of C. viridis species have repeatedly been linked to disturbance in reef environments (e.g. Rützler, 2002;). Bioeroding sponges of the C. viridis complex are believed to be relatively tolerant to environmental deterioration and able to benefit from increased substrate availability after coral mortality (reviewed in Schönberg et al., 2017aSchönberg et al., , 2017b. At our West Indian sample sites reports on reduced reef health largely related to sedimentation (De et al., 2015Hussain et al., 2016;Manikandan et al., 2016). At Carter's (1887) and Dendy's (1916) historical sample sites of C. thomasi sp. nov. pollution and thermal bleaching may be more relevant . Thomas' (1972Thomas' ( , 1979Thomas' ( , 1986 sample sites in Palk Strait have undergone degradation due to coral mining, pollution and bleaching events (Manikandan et al., 2014). The Andaman and Nicobar Islands have been regarded as comparatively unperturbed reef environments, but river sediment discharge into the Bay of Bengal, tsunami damage and global change have taken their toll (e.g. Brown, 2007). We therefore think that monitoring the abundance, distribution and boerosion capacity of dominant C. viridis complex species such as C. thomasi sp. nov. is essential in order to recognise changes in the benthic community over time and to develop suitable strategies for protecting and managing the coral reef ecosystem in the region (Schönberg, 2015).   Dendy (1916) at Okha, in the Gulf of Katchchh, as Cliona coronaria. (6) Reported by Thomas (1972Thomas ( , 1979 and Devi et al. (2011) (2017) found that Cliona laticavicola is conspecific with Cliona delitrix.   spirasters of NIO/BOD/P2 (MGB 21) and NIO/BOD/P4 (MGB 23), respectively, which were predominantly bow-shaped, generally thicker and had shorter, often reduced spines that could merge, or had tylar swellings. (7.4) NIO 776 had more delicate spirasters with more pronounced spine bouquets. From this sponge we also display an example of a branching spiraster (arrow, 2x the enlargement) and of a putatively foreign spiraster with conical spines.