The Bryozoan Margaretta cereoides as Habitat-Former in the Coralligenous of Marzamemi (SE Sicily, Mediterranean Sea)

Abstract

Although several bryozoans are considered habitat-former species, allowing colonisation by epibionts and promoting biodiversity, studies dealt so far with only some, usually constructional, species. In this context, the present study focuses on the epibiosis found on the erect mineralised but flexible species Margaretta cereoides, contributing with some macroalgae to the canopy formation on a coralligenous build-up collected at 36.7 m depth offshore Marzamemi (Ionian Sea, Mediterranean) in the frame of the project CRESCIBLUREEF. Focusing on bryozoans and serpulids we documented high diversity (species richness) on the investigated substrata (M. cereoides, Flabellia petiolata, Peyssonnelia rubra, Osmundaria volubilis and all other fleshy algae including Phyllophora crispa). Diversity, total number of colonies/specimens hosted, and number of colonies/specimens per cm² are higher on M. cereoides than on algae, possibly in relation with the morphology, robustness and persistence of the surface offered by the bryozoan contrasting with the flexible and ephemeral nature of most algae, often provided with chemical defenses. Bryozoan and even more serpulid epibiont diversity on M. cereoides is also higher than those reported for studied species in the Mediterranean and for Flustra foliacea from North Europe.

1. Introduction

Several bryozoans have often been considered relevant habitat-formers able to create 3D structures, enhancing habitat structural complexity and promoting biodiversity (e.g., Wood et al., 2012 [1]; Lombardi et al., 2014 [2]; Lombardi et al., 2020 [3]). Studies have especially been devoted to “frame builders”, i.e., species with robust skeletons developing large-sized colonies and able to produce bioconstructions by growing on each other and/or fusing their branches (see [1]; Cocito, 2004 [4] for a review). In the Mediterranean Sea, data mostly refer to Pentapora fascialis (Pallas, 1766) and Schizoporella errata (Waters, 1878) ([2,4]; Holon & Harmelin, 2013 [5] and references therein). Both species typically develop large colonies of some dm in diameter and height and exceptionally up to 80-100 cm, branched hemispherical or massive with irregular digitations, respectively (Novosel, 2005 [6]; Cocito et al., 2006 [7]; AR personal observations). Colonies often coalesce to form monospecific bioconstructions supporting high biodiversity. Pentapora fascialis from 8–22 m depth in the Ligurian Sea, hosted 84 species, including 27 bryozoans and 2 serpulids (Ferdeghini & Cocito 1999 [8]). A lower number of epibionts (36 species, including 4 bryozoans and 3 serpulids) has been reported for S. errata from 0.3–2 m depth in the same area [8], and even fewer species (28, including 5 bryozoans and 1 serpulid) were observed on colonies from Minorca examined by Maluquer (1985) [9]. High biodiversity figures have also been reported for other species with erect-rigid skeletons developing cm- to dm-sized colonies, i.e., Myriapora truncata (Pallas, 1766), Smittina cervicornis (Pallas, 1766) and Reteporella couchii couchii (Hincks, 1878), supporting a total of 22, 28 and 37 species, respectively (Cocito, 2009 [10]).

However, less attention has been devoted to non-constructional species, though some of them are known to develop dense assemblages of colonies hosting several epibionts [1]. For instance, this is the case for Flustra foliacea (Linnaeus, 1758) from around the British Isles and the North Seas of Europe (e.g., Stebbing, 1971a [11]; Connor et al., 2004 [12]), a poorly calcified species forming large (up to 15 cm high), erect, flexible colonies with irregularly ribbon-like to flabelliform bilaminar fronds. Colonies from 15–20 m off Oxwich Point, South Wales analysed by Stebbing (1971b) [13] supported a total of 42 species, including 1 serpulid and 25 bryozoans. Bitschofsky et al. (2011) [14] reported 42 bryozoan epibionts on this bryozoan after a more extensive study on 51 samples from some North Sea localities, stored in several museums and collected in a wider depth range, in a time span of about 240 years (1776 to 2008).

In the Mediterranean, habitat-forming, non-constructional bryozoans have been unreported so far except for Cellaria salicornioides, a species forming extensive carpets on muddy bottoms at ca. 35 m depth in the Adriatic Sea off Rovinj, Croatia, from where McKinney & Jaklin (2000) [15] listed 54 species, including 31 bryozoans and 4 serpulids. No data are so far available for Margaretta cereoides. This is a cheilostome species forming erect, articulated bush-like colonies, usually up to 10 (rarely 15) cm high, that attach to the substratum by means of chitinous rootlets. Branches consist of heavily calcified, jointed, slender (10–20 mm long and ca. 1 mm in diameter), straight to gently bending internodes arising laterally from the parental ones singularly or in pairs or triplets. Chitinous rootlets mostly develop in basal internodes to form bundles of anchoring rhizoids and occasionally at different heights, allowing the lateral shoring of the colony. Known from the Mediterranean and the Atlantic, M. cereoides is a sciaphilic species thriving in sites with sustained currents in shallow waters (from ca. 5 m down to ca. 50 m depth). In the Mediterranean, it usually occurs in shaded habitats associated with rhizomes of Posidonia oceanica (Linnaeus) Delile meadows, at the entrance of submarine caves, on the underlayer of algal and Cellaria carpets and the precoralligenous (e.g., Gautier, 1962 [16]; Ferdeghini et al., 2000 [17]; McKinney & Jackin, 2001 [18]; Di Martino & Taylor, 2014 [19]; A.R., personal observations). Margaretta cereoides has also been reported as a habitat-forming species developing dense meadows on a rocky plateau at 10 m depth at Lastovo Island, Adriatic Sea [6].

Several colonies of this species were observed associated with coralligenous build-ups offshore of Marzamemi (South of Siracusa, Gulf of Noto, Ionian coast of SE Sicily) during underwater surveys in the frame of the two years long project FISR 04543 “CRESCIBLUREEF”, which aimed at understanding the tempo and mode of the Coralligenous inception and development (see Rosso et al., 2022 [20]). In this area, a Coralligenous habitat was first mapped at the beginning of the 1980s (Di Geronimo in Violanti et al., 1990 [21], figure 3), and discrete build-ups were examined in the last two decades (Di Geronimo et al., 2001 [22]; 2002 [23]; Rosso & Sanfilippo, 2009 [24]). New remote acquisition and Remote Operated Vehicle (ROV) documentation confirmed the occurrence of coralligenous bioconstructions, mostly in the shape of pillars clustered or locally slightly spaced apart, arranged in two main belts approximately between 36 and 100 m in depth, seemingly in relation to the local terraced seascape (Varzi et al., 2023 [25]). Image analysis and computed axial tomography indicated that the internal framework mainly consists of coralline algae and subordinate invertebrates, mostly bryozoans and serpulids producing extensive cavities largely filled with sediment (Bracchi et al., 2022 [26]). As indicated by ROV imaging, the outer surfaces are covered with a dense canopy of soft sciaphilic algae to which subordinate erect bryozoan colonies add at a higher depth. In particular, one of the collected build-ups showed a considerable number of colonies of the bryozoan M. cereoides intermingled with the algae of the canopy. The Chlorophyta Flabellia petiolata (Turra) Nizzamudin and the Rhodophyta Osmundaria volubilis (Linnaeus) R.E. Norris and Peyssonnelia rubra (Greville) J. Agardh were more abundant and obvious. In order to investigate and assess the habitat-forming role of the bryozoan M. cereoides, this paper aims at (1) analysing the epibiont assemblage associated with M. cereoides and its diversity; (2) comparing this epibiont assemblage with those found on algae in the same canopy; (3) discussing the possible differences/affinities of these epibiont assemblages in relation to the nature of their basibionts and (4) comparing the epibiont assemblage found on M. cereoides with those occurring on other habitat-forming bryozoans.

2. Materials and Methods

The studied build-up originates from a field of discrete columnar coralligenous bioconstructions emerging from a biogenic coarse-grained bottom characterized by abundant living maerl rhodoliths in the middle of a submarine channel (see [25,26]). This field was detected in June 2021 during a preliminary survey investigating the Coralligenous distribution in the area and the underwater characterization of the bioconstructions. The build-up (indicated with the letter E) was collected on the 3^rd of August, 2021 during the second field survey of the project CRESCIBLUREEF at 36°43.454′ N; 15°09.657′ E, 36.7 m depth (Figure 1), with a sea bottom temperature of 18 °C, following examination of the ROV_290621_06 below the local distribution of the P. oceanica meadow (Figure 2A).

The build-up consisting of calcareous algae and subordinate encrusting invertebrates (mainly bryozoans, serpulids and sponges) was stout and irregularly fungiform (Figure 2), with a height of 38 cm and a circumference of 52.5 cm at the top, 112 cm at an intermediate height and 71 cm at its base, corresponding approximately to ca. 34, 72 and 50 cm in main cross sections.

The surface was covered with a dense canopy including fleshy algae and colonies of the erect bryozoan M. cereoides. These organisms were removed manually after collection and the photogrammetric acquisition of the build-up in its original condition (see [26] for further details).

For the present investigation, the samples CBR2_4_26, CBR2_4_27, CBR2_4_28 and CBR2_4_30 were examined. They consist of the material collected from the build-up surface including soft algae, bryozoans and a mix of non-constructional organisms found on the carbonate bioconstruction (stored as sample CBR2_4_21c and not considered in the present study) but not specimens/colonies found detached from the structure, even if possibly fallen from it at any time during treatment. Materials from all these samples were preserved in alcohol and grouped for the scope of the present investigation. In the Palaeoecology laboratory of the Department of Biological, Geological and Environmental Sciences (DipBioGeo), University of Catania, large-sized elements forming the canopy were separated and examined for their epibionts focusing on skeletonized taxonomic groups. All material was examined and identified under Zeiss Discovery 8 stereomicroscopes with different magnification configurations. Selected specimens were photo-documented using a photo camera and the software imaging program ZEN 3.1 connected to a stereomicroscope. Several bryozoan colonies were analysed untreated and/or cleaned but uncoated under a Tescan Vega 2 LMU, Low Vacuum Scanning Electron Microscope (SEM) at the Microscopical Laboratory of the DipBioGeo. Images were generated using back-scattered electrons. Macroalgae were identified on the basis of diacritical morpho-anatomical characters using the most recent taxonomic literature. Observations were carried out on thalli preserved in an alcohol–seawater solution. For microscopic observations, some specimens were decalcified with 10% HNO₃ and stained with 1% aqueous aniline blue acidified with dilute HCl to highlight pit-connections, or with Lugol solution to highlight plastids under a Zeiss Axioplan (Göttingen, Germany) microscope. Sections were made with a razor blade under a Zeiss stereomicroscope.

Living and dead assemblages were separated for all detected organisms. Bryozoan colonies were considered alive when they included at least some functional zooids. For serpulids, the occurrence of the worm tissues and the operculum inside the tube was checked. Bryozoan colonies were counted as units when they clearly showed the ancestrulae and/or peri-ancestrular areas. The number of targeted algae of the canopy, i.e., Flabellia petiolata, Osmundaria volubilis and Peyssonnelia rubra, was scored soon after detachment counting separately individual blades, either encrusted or not by epibionts. The green alga F. petiolata consisted of cylindrical stems of 1–3 cm arising from a stoloniferous base formed by unpigmented, irregularly branched, some centimetres long filaments, each bearing a flabellar to rhomboidal thin and extremely flexible blade, usually about 1–2 cm but up to ca. 4 cm wide in the examined material. O. volubilis is a red alga showing erect portions 2–7 cm high in the present samples, consisting of narrow ribbon-like, helicoidally coiled blades fixed to substrate by a discoid base. P. rubra is a red alga consisting of groups of roundish blades 1–3 cm in diameter attached to the substrate by pluricellular rhizoids. For F. petiolata, individual stems and blades were counted separately because several stems missed their blades. Each individual helicoidal and roughly circular blade was counted for O. volubilis and P. rubra, respectively. Further algae were not counted because they were decidedly smaller, or they were represented by more fragmented blades/branches. Colonies of M. cereoides were accurately detached from the build-up trying to avoid separation of their internodes, and their number was also compared to the clusters of basal rootlets observed still attached on the build-up surface when examined for identification of all live and dead constituent organisms (Donato et al., 2022 [27]). In order to provide an indication of the colonisable surfaces offered to epibionts by all constituents of the canopy, we estimated their exposed surfaces following the methods reported below. Branches of M. cereoides and stems of F. petiolata were assimilated to slender cylinders of ca. 1 and 0.5 mm in diameter, respectively, and their length was measured to calculate the outer surface. Blades of F. petiolata, O. volubilis and P. rubra were put on a graph paper, and the covered areas were roughly measured and doubled to consider both sides.

The statistical analysis was performed with Primer 6.1.12 software (Primer-e/Quest Research Limited, Albany, New Zealand) on species (variables) vs. substrates (samples) matrixes. Diversity indexes were calculated with Primer software on the raw number of species per sample.

Material is housed in the Palaeontological Museum of the DipBioGeo, University of Catania (PMC), with bryozoans and serpulids in the CBR Rosso and Sanfilippo collections, respectively. Catalogue numbers are reported for figured species only (see below).

3. Results

The examined build-up was selected because colonies of the bryozoan Margaretta cereoides were obvious in the canopy (Figure 2B) among thalli of both partly calcified algae, such as P. rubra, and soft-bodied species. These latter were mostly represented by F. petiolata and O. volubilis, together with Caulerpa cylindracea Sonder, and less abundant and/or obvious species, including Phyllophora crispa (Hudson) P.S.Dixon, Rytiphlaea tinctoria (Clemente) C. Agardh and Rhodymenia spp. The surface of the build-up resulted from the superimposition of more or less tightly packed laminae of non-geniculate coralline algae, including Lithophyllum stictiforme (Areschoug) Hauck, mineralised Peyssonneliaceans and encrusting sponges [26].

A total of 82 blades of F. petiolata and ca. 40 further stems missing the typical enlarged blade were collected offering an estimated surface of ca. 576 cm² (

Table 1. Summary of bryozoans and serpulids found as epibionts on the engineer species of the canopy in the examined coralligenous build-up E collected offshore Marzamemi (Ionian Sea). NA: Not applicable.

Engineer Species of the Canopy	Margaretta cereoides		Flabellia petiolata		Peyssonnelia rubra		Osmundaria volubilis		Further Fleshy Algae		Total Canopy
Colonies/Thalli	87		82 + 121 Stems		54		20		Several, Not Counted
Approximate Available Surface (cm2)	440		576		217		161		Not Estimated
	L	D	L	D	L	D	L	D	L	D	L	D
Bryozoan number of species	47	9	37	4	19	2	19	1	41	4	63	14
Bryozoan total number of colonies	457	83	324	12	81	2	48	1	220	11	1132	103
Bryozoan colonies/cm2	1.02	0.18	0.56	0.02	0.37	0.01	0.30	0.01	NA	NA	NA	NA
Bryozoan relative species percentages (%)	74.60	14.29	58.73	6.35	30.16	3.17	30.16	1.59	65.08	6.35	100	100
Bryozoan relative colony percentages (%)	40.37	80.58	28.62	11.65	7.16	1.94	4.24	0.97	19.43	10.68	100	100
Serpulid number of species	11	9	8	3	3		3		5	4	12	9
Serpulid total number of specimens	158	161	139	3	19		4		58	7	378	169
Serpulid specimens/cm2	0.36	0.37	0.24	0.01	0.09		0.02		NA	NA	NA	NA
Serpulid relative species percentages (%)	91.67	100	66.67	33.33	25.00		25.00		41.67	44.44	100	100
Serpulid relative specimen percentages (%)	41.80	100	36.77	1.78	5.03		1.06		15.34	4.14	100	100

). P. rubra and O volubilis were less abundant, represented by 54 and 20 blade units and accounting for 217 and 161 cm², respectively (Table 1). Margaretta cereoides occurred with 87 colonies offering a surface of ca. 440 cm² (Table 1).

Besides the bryozoans and serpulids focused on in this paper, several epibionts occurred in the canopy, including abundant flesh algae and mineralised algal species largely represented by Hydrolithaceae (among which Hydrolithon spp.), common foraminifera mostly represented by the slightly attached species Rosalina bradyi (Cushman, 1915) and Lobatula lobatula (Walker & Jacob, 1798) and the encrusting Miniacina miniacea (Pallas, 1766), molluscs (mostly Gregariella semigranata (Reeve, 1858)), and a few brachiopod specimens belonging to Joania cordata (Risso, 1826) and Argyrotheca cuneata (Risso, 1826).

Focusing on bryozoans (Table 1,

Table 2. Relevance (numbers and relative percentages) of the different bryozoan orders (cyclostomes, ctenostomes and cheilostomes) in the epibiont assemblages found on the bryozoan Margaretta cereoides and main algae of the canopy of the build-up E collected offshore Marzamemi (Ionian Sea). L: living and D: dead assemblages.

Engineer Species of the Canopy	Margaretta cereoides		Flabellia petiolata		Peyssonnelia rubra		Osmundaria volubilis		Further Flesy Algae		Total Canopy
Epibiont Bryozoan Species	L	D	L	D	L	D	L	D	L	D	L	D
Total number of cyclostome species	10	2	8	2	3		6		9		11	4
Cyclostome species %	21.3	22.2	21.6	50	15.8		31.6		22		17.5	30.8
Total number of cyclostome colonies	97	4	80	4	10		9		56		251	7
Cyclostome colonies %	21.2	4.82	24.7	33.3	12.3		18.8		25.5		22.2	6.8
Total number of ctenostome species	5		2		2		1		3		5
Ctenostome species %	10.6		5.41		10.5		5.26		7.32		7.94
Total number of ctenostome colonies	38		18		5		1		5		66
Ctenostome colonies %	8.3		5.6		6.2		2.1		2.3		5.8
Total number of cheilostome species	32	7	26	2	14	2	12	1	29	4	47	9
Cheilostome species %	68.1	77.8	70.3	50	73.7	100	63.2	100	70.7	100	74.6	69.2
Total number of cheilostome colonies	322	79	226	8	66	2	38	1	159	11	815	96
Cheilostome colonies %	70.5	95.2	69.8	66.7	81.5	100	79.2	100	72.3	100	72.0	93.2
Total number of bryozoan species	47	9	37	4	19	2	19	1	41	4	63	13
Total number of bryozoan colonies	457	83	324	12	81	2	48	1	220	11	1132	103

and

Table 3. Relevance (numbers and relative percentages) of the different growth morphologies developed by epibiont bryozoans found on the bryozoan Margaretta cereoides and main algae of the canopy of the build-up E collected offshore of Marzamemi (Ionian Sea). Note that morphotypes have been grouped in only three main types for the sake of simplicity.

Engineer Species of the Canopy	Margaretta cereoides		Flabellia petiolata		Peyssonnelia rubra		Osmundaria volubilis		Further Fleshy Algae		Total Canopy
Epibiont Bryozoan Species	L	D	L	D	L	D	L	D	L	D	L	D
Total number of erect flexible species	16		10	1	6		4		11		17	1
Erect flexible species %	34		27	25	31.6		21.1		26.8		27	7.7
Total number of erect flexible colonies	133		72	2	21		10		53		289	2
Erect flexible colonies %	29.1		22.2	16.7	25.9		20.8		24.1		25.5	1.94
Total number of encrusting species	27	9	26	3	12	2	12	1	25	4	41	12
Encrusting species %	57.4	100	70.3	75	63.2	100	63.2	100	61	100	65.1	92.3
Total number of encrusting colonies	285	83	244	10	56	2	33	1	132	11	752	101
Encrusting colonies %	62.4	100	75.3	83.3	69.1	100	68.8	100	60	100	64.4	98.1
Total number of erect rigid species	4		1		1		3		5		5
Erect rigid species %	8.51		2.7		5.26		15.8		12.2		7.94
Total number of erect rigid colonies	39		8		4		5		35		91
Erect rigid colonies %	8.53		2.47		4.94		10.4		15.9		8.1
Diversity (number of species)	47	9	37	4	19	2	19	1	41	4	63	13
Total abundance (number of colonies including undetermined ones)	457	83	324	12	81	2	48	1	220	11	1132	103

; Supplementary Materials Table S1; Figure 3A,B, Figure 4, Figure 5A–G, Figure 6 and Figure 7A–H,J,K) and serpulids (Table 1 and

Table 4. Relevance (numbers and relative percentages) of the two serpulid subfamilies in the epibiont assemblages found on the bryozoan Margaretta cereoides and main algae of the canopy of the build-up E collected offshore of Marzamemi (Ionian Sea). L: living and D: dead assemblages.

Engineer Species of the Canopy	Margaretta cereoides		Flabellia petiolata		Peyssonnelia rubra		Osmundaria volubilis		Further Fleshy Algae		Total Canopy
Epibiont Serpulids	L	D	L	D	L	D	L	D	L	D	L	D
Total number of serpuline species	8	7	5	3	3		3		5	4	9	7
Serpuline species %	72.7	77.8	62.5	100	100		100		100	100	75	77.8
Total number of spirorbine species	3	2	3								3	2
Spirorbid species %	27.3	22.2	37.5								25	22.2
Total number of species	11	9	8	3	3		3		5	4	12	9
Total number of serpuline specimens	129	87	71	3	19		4		57	7	292	95
Serpuline specimens %	81.6	54.0	51.1	100	100		100		98.3	100	77.2	56.2
Total number of spirorbine specimens	29	74	68						1		86	74
Spirorbine specimens %	18.4	46.0	48.9						1.72		22.8	43.8
Total number of specimens	158	161	139	3	19		4		58	7	378	169

; Supplementary Materials Table S2; Figure 3C,D, Figure 5A,B,H–M and Figure 7I,K), the total species richness was of 63 and 12 species, respectively. When examining the different substrata separately, M. cereoides alone hosted 47 bryozoans and 11 serpulids accounting for ca. 75% and 92% of the total diversity, respectively.

Values were also elevated but lower for F. petiolata (59% and 67% for bryozoans and serpulids, respectively) and all soft algae considered together (65% and 42%), but decidedly lower for both P. rubra and O. volubilis (both accounting for the 30% and 25%, respectively). Figures for dead assemblages were lower but with roughly similar trends. The highest values were found for bryozoans and serpulids on M. cereoides, with 9 species each accounting for the 64% and 100% of the dead bryozoan and serpulid assemblages, respectively. Intermediate values were observed for F. petiolata (29% and 33%) and all algae together (29% and 44%), and the lowest values for P. rubra and O. volubilis, with bryozoans represented by only 2 (14%) and 1 (7%) species, respectively, whereas serpulids were missing.

A total of 1132 living and 103 dead bryozoan colonies were found (Table 1, Table 2, Table 3 and Table S1). Most colonies (457 living + 83 dead, corresponding to 40.4% and 80.6% of all colonies found on the canopy of the E build-up) encrusted M. cereoides. Further 324 living and 12 dead colonies (corresponding to 28.6% and 11.6%, respectively) occurred on F. petiolata, and 220 living and 11 dead colonies (19.4% and 10.7%, respectively) on further fleshy algae grouped together. Figures were lower on both P. rubra and O. volubilis, showing 81 and 48 living colonies and related relative percentages of 7.2% and 4.2%, respectively. Only 2 and 1 dead colonies occurred on P. rubra (1.9%) and on O. volubilis (1%). Margaretta cereoides also showed the highest colonization values per surface unit (slightly more than 1 live colony/cm² and 0.2 dead colonies/cm²) in comparison with F. petiolata, P. rubra and O. volubilis, which showed colonization values of 0.56, 0.37 and 0.30 live colonies/cm² but negligible values for dead colonies.

Living colonies belonged to 63 species including 11 cyclostomes, 5 ctenostomes and 47 cheilostomes, accounting for 17.5, 7.9 and 75.0%, respectively (Table 2). Relationships between these taxonomic groups remained very similar when the number of colonies was considered, with cheilostomes still accounting for 72% but with a slight decrease of ctenostomes (5.8%) counterbalanced by an increase of cyclostomes (22.2%). However, both species richness and colony abundance varied if the epibionts associations found on M. cereoides and the different algae were considered separately. The total number of living species ranged from 19 (on both P. rubra and O. volubilis) to 47 (on the bryozoan), with an intermediate species richness found on F. petiolata (37 species) and further fleshy algal species considered together (41 species).

Analogously, the total number of colonies varied strongly. Margaretta cereoides showed the highest values with 457 colonies accounting for 40.4% of the total association, followed by F. petiolata (324 colonies = 28.6%) and all other algae (220 colonies = 19.4%), whereas P. rubra and O. volubilis hosted only 81 and 48 colonies, representing 7.2% and 4.2%, respectively. Some variability was also observed between the three orders (cyclostomes, ctenostomes and cheilostomes) on the different types of substrata of the canopy, with ranges of ca. 15–31%, 5–10% and 63–74% at species and of ca. 12–25%, 2–8% and 69–81% at colony level, respectively. These figures indicate comparable dominances of cheilostomes at both species and colony level in assemblages on individual substrata and lower relative percentages (but wider variabilities) for ctenostomes (and especially for cyclostomes) as regards colony abundance relative to species richness (Figure 4).

Encrusters dominated with 41 species and 752 colonies (Table 3), invariably representing about 65% of the total assemblage, followed by erect flexible species (17 species and 289 colonies = ca. 27%), whereas erect rigid morphotypes were decidedly subordinate (5 species and 91 colonies representing 7.9% and 8.1%, respectively). Leaving out the epibiosis found on soft-bodied algae considered together, encrusting species were dominant on both M. cereoides and F. petiolata (27 and 26 species, respectively), developing 285 colonies (62% of the assemblage) on M. cereoides and 244 colonies (75% of the assemblage) on F. petiolata. In contrast, both P. rubra and O volubilis hosted 12 species (63%), represented by 56 (69%) and 33 (69%) colonies, respectively.

Ranges of variability were slightly lower for rigid erect morphotypes when species were considered (3–15%), but wider for erect flexible and encrusting species that showed variations of 13% at species level and of 16–18% at colony level. The highest relative percentages of encrusters occurred on F. petiolata (see above), whereas erect flexible species were more diversified and abundant (16 species = 34% and 133 colonies = ca. 29%) on M. cereoides.

Only ca. 1/3 of the species were represented by more than 10 colonies, with only 9 species ranging from 24 to 48 colonies and 4 species exceeding 50 colonies, whereas 18 species occasionally occurred with only 1–2 colonies (Supplementary Materials Table S2). With 228 colonies, the encrusting (celleporifom) Celleporina caminata (Figure 5A,C and Figure 7J,K) dominated (mostly on M. cereoides and subordinately on the stems of F. petiolata), usually developing small (1–2 mm) pisiform colonies. The erect flexible species Scrupocellaria delilii (72 colonies) (Figure 7C) was also abundant on M. cereoides (38 colonies) and subordinately on F. petiolata (10 colonies) and further soft-bodied algae (16 colonies). The third most abundant species was Mecynoecia delicatula (64 colonies) developing small erect rigid colonies usually consisting of an extremely reduced encrusting part, often divided in 2–3 (rarely 4) branches, each giving rise to short (less than 1 cm long), erect, often fertile portions (Figure 5B and Figure 6A). This species was common (34 colonies) on M. cereoides but subordinate on F. petiolata (8 colonies) and on other soft-bodied algae (16 colonies), P. rubra (4 colonies) and O. volubilis (2 colonies). Aetea cf. truncata developed relatively common extensive networks on almost all elements of the canopy (but subordinately on P. rubra and O. volubilis), from where some short branches sporadically developed. Further relatively abundant species were the encrusters Copidozoum planum (48 colonies largely on F. petiolata and further soft-bodied algae), usually forming small sized colonies, Annectocyma major (42 colonies) and Beania mediterranea (40 colonies, mostly on M. cereoides), as well as the erect flexible Beania cylindrica (Figure 5G and Figure 6F) (42 colonies, especially on M. cereoides and P. rubra). Both Beania species were relatively larger, with B. cylindrica developing erect branches up to 2 cm long. Further common species were the encrusters Chorizopora brongnartii (Figure 5D and Figure 7G) (29 colonies, usually extremely small-sized, especially found on the upward facing surface of P. rubra) and Turbicellepora coronopus (26 colonies, especially on M. cereoides) and the erect flexible species Crisia ramosa (35 colonies), Amathia delicatula (34 colonies, especially on M. cereoides) and Cabera boryi (24 colonies, especially on F. petiolata).

The highest number of species with living colonies (48) occurred on M. cereoides, and most of them (29) were shared with F. petiolata and other constituents of the canopy. Further 7 species were shared with other algae, and 11 were only found on M. cereoides. Among them Bugulina cf. fulva (Figure 6C), with 6 colonies, was the most abundant. In contrast, out of the 18 species never found on M. cereoides, 4 occurred only on F. petiolata, each with 1 to 4 colonies.

When looking at dead bryozoans still attached on their substrata, the total number of species (14) and colonies (103) was remarkably low, but two species (Schizobrachiella sanguinea and Plesiocleidochasma mediterraneum) were added that were not present in the living population. A single species (C. caminata) with 72 colonies accounts for more than 70% of the dead assemblage, whereas all other species were present with 1–3 colonies, except for Turbicellepora coronopus with 5 colonies. Most species and colonies were found on M. cereoides (78% and 76% of the dead assemblage, respectively) and subordinately on the stems of F. petiolata or on the thalli of the relatively robust Phyllophora crispa. In all instances, however, the number of colonies per cm² was almost negligible.

Serpulids occurred with 378 living specimens and 169 empty tubes (Table 1 and Table 4; Supplementary Materials Table S2). Like for bryozoans, the highest number of specimens (158 living and 161 dead) occurred on M. cereoides, accounting for 41.8% and 95.3% of the living and dead assemblages, respectively. Further 139 living and 3 dead specimens (36.8% and 1.8%, respectively) encrusted F. petiolata and 58 living and 7 dead specimens (15.3% and 4.1%, respectively) were found on other fleshy algae considered together. Only few living specimens (but no dead ones) occurred on P. rubra and O. volubilis, accounting for 5% and 1.1% of the observed assemblage, respectively.

Serpulids were absent on several M. cereoides colonies and algal thalli, and the number of specimens per cm² reached 0.36 on M. cereoides but lowered remarkably to 0.24 on F. petiolata. Figures were quite negligible on P. rubra (0.09) and O. volubilis (0.02). Serpulines decidedly prevailed at both species (73%) and specimen (82%) level with values ranging from 63 to 100% at species level and from 51 to 98% at specimen level. Serpulines also strongly prevailed on each considered constituent of the canopy, except on blades of F. petiolata, where spirorbines reached 68 specimens, accounting for 49% of the whole population.

The most abundant species was the serpuline Josephella marenzelleri (Figure 5I), which occurred with 144 specimens, representing ca. 38% of the assemblage as a whole and 37% on M. cereoides (59 specimens) (Supplementary Materials Table S2). Specimens of this species possessing a characteristic minute flexible tube were usually clustered. In contrast, specimens of Spirobranchus triqueter (Figure 5L and Figure 7K), a serpuline species having large sized heavily mineralised tubes, were often found coiled around the internodes of M. cereoides and the stems of F. petiolata. This species occurred with 72 specimens in the whole assemblage (19%), 50 of which were on M. cereoides (32%). The third most abundant species (50 specimens) was the spirorbine Janua heterostropha, representing only 13% of the whole assemblage, but more than 35% on F. petiolata, on which blades this minute spirorbine was particularly abundant, sometimes with clustered juvenile specimens. Analogously, tubes of Filograna sp. (Figure 5J) were mostly found on F. petiolata, though dead specimens were also abundant on M. cereoides. A further spirorbine, Simplaria pseudomilitaris (Figure 5L,M), occurred with 20 specimens (5%), mostly on M. cereoides (16 = 10%). Among further species only the spirorbine Neodexiospira pseudocorrugata (Figure 5K and Figure 7I) and the serpuline Serpula vermicularis (Figure 5H) were relatively abundant, with a total of 16 and 10 specimens, respectively, accounting for 4% and less than 3%. Like the congener, N. pseudocorrugata was especially found (15 specimens) on blades of F. petiolata, where it accounted for ca. 11%.

Among the 12 living species, all but one occurred on M. cereoides and 8 on F. petiolata, whereas only 3 species were found on both P. rubra and O. volubilis.

Dead serpulid assemblage was relatively scant with 9 species and 169 specimens, almost all (158) found on the colonies of M. cereoides. Few dead specimens only sporadically occurred on F. petiolata and other fleshy algae, such as Phyllophora crispa and Rhodymenia sp. All species except Filogranula calyculata (only represented by one dead specimen) also occurred in the living assemblages, though sometimes on different elements of the canopy (see Filograna sp.). The most represented species was S. pseudomilitaris (42% of the serpulid dead assemblage), followed by S. triqueter and Filograna sp. (20% and 17%), whereas J. marenzelleri was subordinate (slightly more than 10%) and all other species rare.

4. Discussion

4.1. Diversity

Almost all bryozoan and serpulid species reported on M. cereoides are typical inhabitants of shelf habitats in the Mediterranean, including some newly described species, such as Microporella modesta (Di Martino et al., 2020) (Figure 7H) and M. bicollaris (Di Martino & Rosso, 2021) (Figure 5E), except for Protula sp., whose taxonomic status needs some investigation. Furthermore, specimens of Serpula vermicularis show a peculiar ornamentation consisting of spines on the central keel, making them similar to those already found in submarine caves from the eastern Mediterranean and identified as S. vermicularis forma echinata (Sanfilippo et al., 2017a [28]). This finding confirms the occurrence of a morphologically distinguishable population of S. vermicularis, whose distribution and taxonomic status deserve further investigation.

Unlike algae occurring with at least two non-indigenous species (C. cylindracea and C. taxifolia v. distichophylla), both bryozoans and serpulids are only represented by species considered native to the basin. Both bryozoans and serpulids occurring on M. cereoides represent a relevant part of the total number of species so far known from the Mediterranean for these two taxonomic groups. Bryozoans on the examined M. cereoides colonies and the whole build-up represented 8.5% and more than 11%, respectively, of the 556 bryozoan species reported by Rosso and Di Martino (2016) [29]. Analogously, serpulids accounted for the 16% and more than 17% of the 69 species obtained summing up those reported by Castelli et al. (2008) [30] for the western basin and by Ben-Eliahu and Fiege (1996) [31] and Çinar et al. (2006, 2021) [32,33] for the eastern Mediterranean.

Bryozoans found on M. cereoides also account for the 21.9% of the total diversity so far known for the coralligenous habitat (219 species reported in [29]), a figure relatively close to the 28.8% accounted to by the whole assemblage of the canopy in the examined build-up. Serpulids are usually merged within polychaetes in literature and summaries are not available for them. However, the 11 species found only on M. cereoides outnumber the 6 and 10 species reported by Casellato and Stefanon (2008) [34] and by Laubier (1966) [35] for the Adriatic Sea and the Marseille area, respectively, and parallel the 11 species reported from 90-100 m deep coralligenous assemblages in the Gulf of Noto by [24].

The Margalef index (d = 10.4) also indicates a high diversification for bryozoan and serpulid living assemblages (

Table 5. Margalef (d), Pielou (J’) and Shannon (H’) indexes for bryozoan and serpulid living assemblages found on the bryozoan Margaretta cereoides and on algae of the canopy of the build-up E collected offshore Marzamemi (Ionian Sea).

	M. cereoides	F. petiolata	P. rubra	O. volubilis	Further Algae	All Canopy
d	8.979201569	7.234201966	4.633102932	5.424673103	8.568186705	10.39512588
J’	0.727544638	0.844863311	0.869820647	0.942942333	0.828398275	0.771925758
H’	2.990844237	3.216109475	2.688652546	2.914674783	3.223977221	3.363055763

). Figures are also elevated for M. cereoides and F. petiolata considered separately (8.9 and 7.2, respectively) and for further algae considered together (8.5), but lower at 5.4 and 4.6 for O. volubilis and P. rubra, respectively, although still indicating a good species richness.

The high values of both J’ (ranging from 0.7 to 0.9) and H’ (ranging from 2.6 to 3.3) indexes indicate a relatively equilibrate distribution of epibionts on available substrata.

4.2. Habitat-Forming Role of M. cereoides and Algae in the Canopy of the Studied Coralligenous Build-Up

Data reported above clearly points to the relevance of M. cereoides as a substratum, at least for bryozoans and serpulids. This relevance is also supported by the high figures in diversity (species richness), total number of colonies/specimens hosted and number of colonies/specimens per cm² observed on M. cereoides in comparison with the algal constituents found in the canopy developed on the E build-up collected off Marzamemi in the Ionian Sea.

Comparisons of the epibiosis developed on algal and animal substrata, especially bryozoans, are missing in the literature. However, in an investigation aimed to highlight variation in species richness of epifaunal bryozoan assemblages and their composition in relation with different algal functional groups and related structural complexity on an array of filamentous to large laminar (Laminariales including Macrocystis) algae of SW Atlantic, Liuzzi and Lopez Gappa (2010) [36] found no significant differences except between filamentous and foliose algae. Analogously, Hayward (1980) [37] found that besides a small number of primary epibiont species, i.e., species directly depending on the host algal substratum as a resource and hence establishing species-specific and even obligate associations with their hosts, other epiphytes (which he indicated as “secondary epiphytes”) simply are opportunistic eurytopic species settling on any kind of available algal substrate. Hayward [37], however, suggested that epibionts settle on particular plant species in relation to particular physical factors, including the size of the colonisable surface and its texture and rigidity. Differences in the colonization pattern between laminar corallinales and fleshy algae has also been remarked by Gallardo et al. (2021) [38] for macroalgae collected at 15 m depth off Mataró, NE Spain. These authors assessed that the nature (soft vs. hard consistency of the thallus) and the architecture (branched or crusted and convoluted) of macroalgae can limit the size and abundance of the epibiont invertebrates and can condition the populations of organisms that settle on them.

Comparisons of the present situation with data from the literature are difficult due to the different investigated geographical and ecological settings, the taxonomic and even ecological and functional groups dealt with and the approaches used. After examination of museum collections of algae originating from a wide geographical area, for instance, [36] found a remarkably low total diversity of 36 epibiont bryozoans on 36 encrusted algal genera.

Notwithstanding the relatively low differentiation of algal basibiont morphologies and the single sample and site, in the examined situation, epibiont species richness was comparably higher (63 bryozoans) and differences between substrata more remarkable. This is also indicated by the statistical analysis (Figure 8) that groups more tightly P. rubra and O. volubilis, the two substrata with the lower relevance of both bryozoan and serpulid epibionts. This comparably low occurrence of epibionts could be related to the possible presence of antifouling substances produced by these algae, discouraging epibiont settlement (see Ryland, 1959 [39]; Steinberg and de Nys, 2002 [40]).

Though generalisation is difficult, red and brown algae have been reported as the most chemically prolific (da Gama et al., 2014 [41]). P. rubra was one of the species missing epibionts in all habitats from where it was collected in a study by Uriz et al. (1991) [42]. The particular texture of both surfaces and mostly of the downward facing one, characterized by a mat of rhizoids and entangled sediment, could be unsuitable/unpleasant for the settlement of epibiont species. Indeed, only 1/3 of the bryozoan species and ¼ of the serpulid species, representing ca. 11% and 6% of the total bryozoan and serpulid living assemblages, respectively, were found on these two algae. The number of colonies/specimens was very low for bryozoans and negligible for serpulids, with entire fronds not fouled at all. Species were usually occasional and only some of the most abundant were relatively common except for C brongniartii, a slightly mineralised, encrusting bryozoan that was preferentially found (more than 50% of its colonies) on the upward-facing surfaces of P. rubra. The location of this species, often found on the upward-facing surfaces of algal substrata, could be related with its relative tolerance to light, also indicated by its occurrence/abundance at shallow depths and on buoyant and drifted substrata (e.g., [16]; Koćak and Aydin-Önen, 2014 [43]; Dick and Grischenko, 2017 [44]; Subías-Baratau et al., 2022 [45]; A.R., personal observations). In contrast, F. petiolata, a green alga that resulted as non-toxic in all seasons in a study on antifouling algal toxicity of Mediterranean species by Martì et al. (2004) [46], hosted a comparably higher number of species (37 bryozoans and 8 serpulids) and colonies/specimens (324 bryozoans and 139 serpulids). An amount of 41 bryozoan and 5 serpulid species (with 220 colonies and 58 specimens) were present on further algae, largely represented by fronds of Phyllophora crispa. Whereas no data is available for epibionts on F. petiolata, extensive surveys on epibionts growing on P. crispa from several samples collected in different sites and seasons around the Giglio Island (Tuscany Archipelago, NE Tyrrhenian Sea) allowed (Rossbach et al., 2022 [47]) to recognize a number of bryozoan species belonging to 43 different families and ([47] and Rossbach et al., 2021 [48]) to list 9 serpulid taxa, roughly comparable to figures observed in the present study. As far as we know, together with the record of 54 bryozoan epibiont species on 100 fronds of the large-sized Laminaria rodriguezii Bornet, 1888 (colonisable surface of ca. 10 m²) from ca. 60 m depth at the Apollo Bank (Ustica, S Tyrrhenian Sea), these represent examples of the highest bryozoan diversity associated with algae documented from the Mediterranean.

The highest number of bryozoan and serpulid species, along with the highest number of colonies/specimens per cm², was observed on M. cereoides rather than on other elements of the canopy. This points to the habitat-forming relevance of this bryozoan in the studied build-up, in agreement with the possible “bioconstructional” role suggested for populations of this species in relatively deep, high energy settings of the Columbian Caribbean by Yepes-Narváez (2020) [49]. Margaretta cereoides alone hosts almost 75% and 92% of bryozoan and serpulid species diversity, respectively, providing suitable colonisable surface for most of the detected species, some missing on other substrata.

Algal thalli are fleshy and flexible. This implies that epibionts must be specifically adapted to cope with these characteristics, exhibiting small sizes that can be functional to avoid early detachment. As reported by [37], a weak or incomplete calcification can also help to overcome substrate flexibility, with particular species selecting fronds of the most suitable algal substrates and showing particular distribution patterns on them. Besides unmineralised ctenostomes, in the present instance, we observed some cheilostome bryozoans, such as Escharoides mamillata and further species like Prenantia cheilostoma, which consisted of only few zooids often lacking basal calcification. Analogously, the small spirorbines Janua heterostropha and N. pseudocorrugata, characterized by reduced or missing basal calcification, were nearly exclusively found (sometimes clustered) on the extremely flexible blades of F. petiolata and less frequently on other algae. Tiny and flexible tubes also allowed the success of the serpulid J. marenzelleri not only on F. petiolata but also on further algae, including C. cylindracea (Figure 5I). Similar adaptations were recorded for bryozoans and serpulid assemblages associated with fleshy algae at 5–25 m depth in the Ciclopi Island Marine Protected Area ([28]; Rosso et al., 2019 [50]). The absence/reduction of basal calcification could also allow further functions. Manríques and Cancino (1996) [51] demonstrated that it favors the uptake of algal exudates by particular epibiont bryozoan species preventing zooid degeneration, and Muñoz et al. (1991) [52] reported on the interchange of metabolic substances between the host and the epibionts. This adaptation is also possibly useful for reducing the impact of depressing photosynthetic activity, enhancing the durability of the ephemeral algal fronds and delaying/avoiding the consequent frond loss (Taylor and Wilson, 2003 [53] and references therein), an adverse event for epibionts. Competition for space seems relatively low to absent on algal fronds, probably because of their rapid degradation and turnover constantly providing new colonisable surfaces.

In contrast, the preference for M. cereoides seems related to the different nature of the colonisable surface offered by the bryozoan, i.e., a more consistent (harder) and relatively more persistent surface, in comparison with the algal species usually lasting for one year or less (see McKinney and Jackson, 1989 [54]). Indeed, even if data relating to life histories of all species considered in this study are lacking, we can expect these algae to be more ephemeral than M. cereoides whose colonies, presumably growing for several years, include mineralised internodes thickened by secondary calcification during aging, with basal internodes usually more heavily covered by epibionts. Algae probably have shorter life histories although some parts, and especially the stems of F. petiolata, often persist for some time after degradation/consumption of the blades, as observed in the present instance. Internodes of M. cereoides and stems of F. petiolata share a slender cylindrical, i.e., filiform, shape offering epibionts a 3D-space instead of the 2D-space provided by ribbon-like to laminar algal blades.

It is also worth noting that some of the largest bryozoan colonies and most of the largest serpulid tubes were found on M. cereoides. These include the encrusting unilaminar bryozoans Arthropoma cecilii (Figure 5F and Figure 6E), Schizomavella (S.) linearis (Figure 6D), Copidozoum planum, Plagioecia patina (Figure 6B) and Hippoporina pertusa, as well as the encrusting unlayered celleporiform species C. caminata (Figure 5A,C and Figure 7J) and Turbicellepora coronopus, and also some serpulids, especially S. triqueter (Figure 5L) and S. vermicularis and the elevated, heavily skeletonized spirorbine Simplaria pseudomilitaris (Figure 5M). In addition, large, erect and flexible colonies, such as Bugulina cf. fulva (Figure 6C), preferred Margaretta internodes to settle on. Instead, until present, M. cereoides has only occasionally been reported as a substratum for a few bryozoan species, i.e., the celleporids Turbicellepora robusta (Lopez de la Cuadra and Garcia Gomez, 2001 [55]) and C. caminata [6], this latter species being the most abundant in the present sample.

Whatever the exploited substratum, epibionts are often represented by young colonies/specimens. This could be partly explained by the sampling season (August), a period when almost all species are fertile in the Mediterranean Sea (e.g., [16] for bryozoans and Kupriyanova et al., 2001 [56] for serpulids), and it is possible to find recently settled generations at their early developmental stages and as juveniles, often growing on newly produced algal blades. Some of the detected serpulids, such as S. triqueter, S. vermicularis and spirorbine species, are known as pioneer species (Fava et al., 2016 [57]). With some exceptions, especially on M. cereoides basal articles and some F. petiolata stems and more occasionally on some algae (Figure 5B,J,L), colonies/specimens are often largely spaced from each other, leaving uncovered large available surfaces (cf. densities of colonies/specimens per square centimeter on different substrata in Table 1). Though small and presumably young, however, bryozoan colonies were often fertile, including gonozooids and ooecia, sometimes in zooids not far from the ancestrula in encrusting cheilostomes. Early fertility has often been considered as typical of r-strategist species with opportunistic life histories as an adaptation for exploiting unstable habitats, with ephemeral substrata and/or reduced to unpredictable resource supply ([54]; Taylor, 2020 [58]; Harmelin and Rosso, in press [59]) and already reported for assemblages colonising infralittoral algal communities (e.g., [50]).

The number of dead bryozoan species and colonies, as well as the proportionally higher serpulid species and specimens, was low in comparison to the living assemblage, and only two bryozoan and no serpulid species were added in the thanatocoenoses (dead assemblages). This parallels what usually occurs on hard surfaces of cave walls (e.g., [19]) but contrasts with expectation due to the usual increase in both species and colony/specimen numbers produced by the accumulation of skeletal remains on sedimentary bottoms (e.g., Kidwell and Bosence, 1991 [60] and Rosso, 1996 [61] for bryozoans). However, on a coralligenous build-up, the canopy largely consisted of ephemeral (often short living) substrata, and even “perennial” algae usually lose their fronds during fall or undergo to periodic frond consumption and/or degeneration. The highest number of dead bryozoan colonies and especially of serpulid specimens was observed on Margaretta and the stems of Flabellia, which often persist after the decay of the fan-shaped fronds. This high number of epibionts and especially of dead colonies/specimens on such substrata appears correlated with both their hardness and longevity, sometimes encompassing/exceeding the epibiont’s life span. The relatively low number of observed dead colonies/specimens could also result from their ease detachment from the substratum, possibly favoured by both the smallness of the adhesion surface after the reduction of basal mineralisation (see above) and the possible pellicular shedding of algal epithallial cells, a strategy adopted by some algae to avoid epibiont overgrowth (see Nylund and Pavia, 2005 [62]). These detached colonies/specimens could contribute carbonate bioclasts to the sediments accumulating within crevices and cavities in the coralligenous build-up and in the surrounding soft bottoms. This process also affects Margaretta colonies and internodes that can be detached from their substratum after death and deterioration of chitinous joints and rootlets, with internodes accumulating in the neighboring biodetritic bottoms, as observed in samples collected in the northern sector of the Gulf of Noto at 45 m depth, possibly close to coralligenous build-ups in the area ([61]; Rosso, 1989 [63]).

4.3. Bryozoan and Serpulid Assemblages on M. cereoides and other Large Habitat-Forming Bryozoans

Margaretta cereoides has been reported at least once as a habitat-forming species at 10 m depth in central Adriatic Sea [6], and its “bioconstructional” role has been remarked by [49] in the Columbian Caribbean. However, no study has so far been performed on the associated epibiont community in these areas or elsewhere, and only occasional information is available about particular epibiont species on colonies from the Balearic and the Adriatic seas ([6,55]: see above). Therefore, the present study is the first thorough analysis of the epibiont community associated with M. cereoides, indicating that this species hosts a remarkable diversity.

Indeed, few studies have been so far devoted at investigating epibiont communities on habitat-forming bryozoans in general, unlike for other taxa, especially P. oceanica (e.g., Mazzella et al., 1989 [64]; Nesti et al., 2009 [65]; Di Martino and Taylor, 2014 [66] for a review about bryozoan epibionts), some fleshy algae (e.g., [47,50]; Di Geronimo et al., 1990 [67]) and reef-forming organisms, including coralligenous algae and mesophotic benthic animals (e.g., [24] Harmelin, 2017 [68]; Giampaoletti et al., 2020 [69]; Pica et al., 2022 [70], focusing particularly on bryozoans). In European seas, the most investigated species is possibly F. foliacea. Almost 60 epibiont species have been reported from the northern seas [13,14] on this laminar weakly calcified bryozoan, which grows for up to 10–12 years, developing large colonies usually 10–15 cm high (Ryland, 1976) [71]. Bryozoans on F. foliacea represent ca. 2/3 of the epibionts and largely consist of weakly calcified and articulated species (Supplementary Materials Table S3). It is suggested that this prevalence could be related to both the flexible nature of the colonies (possibly also enhanced by seasonal growth discontinuities) and the species distribution in habitats with moderately strong to strong energy (Hiscock, 1983) [72]. In contrast, studied Mediterranean habitat-forming bryozoans mostly show erect, rigid colonies and are often associated with the Coralligenous. However, only data about total biodiversity, ranging from 22 to 37 species, are available for some species, i.e., M. truncata, S. cervicornis, A. calveti and R. couchii couchii ([10]: see introduction). Lists of bryozoan species (summarized in Supplementary Materials Table S3) are only available for S. errata and P. fascialis, supporting 28-36 and 84 epibiont species, respectively. In these last cases [8,9], bryozoans account for a large proportion of epibiont diversity, with a total of 7 species found on S. errata and 26 species on P. fascialis (ca. 1/3 of the assemblage). Serpulids occurring with 4 and 2 species, respectively, seem to be proportionately more diversified, taking into consideration that their total number in the Mediterranean is comparably very low (see above).

The only flexible bryozoan species so far analysed for epibionts in the Mediterranean is C. salicornioides, sharing with M. cereoides the cellariiform growth morphology of the colonies that consists of mineralised segments articulated by chitinous joints. However, colonies of C. salicornioides are smaller and branches are slender and straight, and they bifurcate only occasionally, producing a single distal branch but not lateral branches in a whorl pattern. McKinney and Jaklin (2000) [15] listed 54 epibiont species on C. salicornioides, forming dense meadows at ca. 35 m depth in the Adriatic Sea. Epibiont assemblage largely consisted of bryozoans (31 species) and subordinately by annelids (7 species, including 4 serpulids). McKinney and Jaklin (2000) [15] also noted a pattern of distribution of epibionts with the highest cover on basal segments and a spatial partition between erect flexible, encrusting and erect rigid colonies dominating on distal, intermediate and proximal (deep) segments, respectively. Though not quantitatively investigated, the number of species and colonies, as well as their coverage, was higher on proximal (oldest) segments on the examined M. cereoides colonies.

If we compare bryozoan and serpulid species richness on these habitat-forming bryozoan species, it appears that M. cereoides hosts a relevant biodiversity (Supplementary Materials Table S3). With 11 species, serpulids decidedly outnumber the 1-4 species reported for other substrata. Some species, i.e., S. vermicularis, Spirobranchus triqueter (usually reported as Pomatoceros triqueter) (e.g., [8]) and S. polytrema, were shared. In contrast, Filograna implexa and Hydroides dianthus were missing on M. cereoides. However, F. implexa could possibly be the same species here reported as Filograna sp., and the absence of H. dianthus could be related to its shallower depth range (Bianchi and Morri, 2011 [73]; Zibrowius, 1973 [74]; Ben-Eliahu and ten Hove, 2011 [75]).

Bryozoans on M. cereoides accounted for 48 species, exceeding the 31 species reported on C. salicornioides, making it the most diverse basibiont for bryozoans in the Mediterranean, also outnumbering the 42 species reported for F. foliacea from the northern seas by [14]. Despite some differences, as many as 23 species were shared between M. cereoides and further Mediterranean basibiont bryozoans, but only 5 of them were also present on F. foliacea. Three of these species are cyclostomes (Diplosolen obelius, Disporella hispida, P. patina) whereas only two, i.e., Schizomavella (S.) linearis and S. scrupea, are cheilostomes (out of the 32 cheilostome species reported). A further two species (Celleporina hassalli and C. brongniartii) have been reported for F. foliacea but not for Mediterranean basibionts other than Margaretta from off Marzamemi. The large divergence in the composition of the epibiont community on F. foliacea in comparison with bryozoan habitat-forming species from the Mediterranean was expected due to the different geographical provenance areas. At the present state of knowledge, however, and due to the scant number of studies, it is difficult to understand if the species found so far that are associated only with a particular basibiont are actually actively selective or if their occurrence is purely occasional. In contrast, the shared taxa could be more eurytopic, as observed for bryozoan species epibiont on algal and plant species (see [36]).