Spatial distribution of the black coral Leiopathes glaberrima (Esper, 1788) (Antipatharia: Leiopathidae) in the Mediterranean: a prerequisite for protection of Vulnerable Marine Ecosystems (VMEs)

Abstract Leiopathes glaberrima is an important component of the deep coral forests of the Mediterranean Sea enhancing benthic biodiversity and forming an essential habitat for many marine organisms. It is a sessile species with high longevity; it is therefore highly affected by fishing impacts and has been listed as a representative taxon of Vulnerable Marine Ecoystems (VMEs). Since information on spatial distribution of L. glaberrima is still scanty and fragmented, an updated synopsis of its distribution in the Mediterranean is given, including two new records in the Strait of Sicily. On the basis of this updated information, hotspots of black coral off the Carloforte Island in the Sardinian Sea, on the Marco Bank and off the Pontine Islands in the Southern Tyrrhenian Sea, and on the Malta Graben in the Strait of Sicily are proposed for implementation of Fishery Restricted Areas.


Introduction
Antipatharians, or black corals, have an important structuring role both in tropical reefs and in temperate ecosystems.
In the last few decades, the coral communities of the Mediterranean Sea have been widely studied. In particular, remotely operated vehicle (ROV) surveys reported that antipatharians living in the outer shelf and upper slope may be locally very abundant (Fabri et al. 2014;Bo et al. 2015).
Leiopathes glaberrima (Esper, 1792), a black coral belonging to the family Leiopathidae, is a major component of the deep coral forests of the Mediterranean Sea defined as "Coral Garden" habitats. Due to the arborescent shape of colonies, forming a complex three-dimensional structure, L. glaberrima is a "habitat-forming" species that increases benthic biodiversity, and serves as feeding, shelter and spawning/ nursery grounds for many associated species (Cau et al. 2013(Cau et al. , 2017Bo et al. 2015).
Leiopathes glaberrima has a wide spatial distribution, occurring in the Pacific Ocean, in the Eastern North Atlantic Ocean (Macaronesian Archipelago and Bay of Biscay) and in the Mediterranean Sea (Opresko & Baron-Szabo 2001). This species occurs in a wide bathymetric range, living between 100 and 2048 m (Molodstova 2014). No data on L. glaberrima larval dispersal are available; however, Miller (1998) found that larvae of Antipathes fiordensis Grange, 1990, a black coral living off New Zealand and belonging to the family Antipathidae, settle close to parent colonies. Although zooplankton is the predominant prey in black corals' diet (Wagner et al. 2012), feeding based on particulate organic matter (POM) was suggested by Carlier et al. (2009). Indeed, the black corals play an important role in the transfer of energy and biomass from pelagic to benthic environments, by recycling POM sinking from the upper photosynthetic layers (Gili & Coma 1998). Since black corals are suspension feeders, their habitat selection is forced mainly by the hydrodynamic regime and the type/availability of substrate (Bo et al. 2009(Bo et al. , 2011a(Bo et al. ,b,c, 2012Wagner et al. 2012). Particularly, L. glaberrima occurs on exposed rocks and bench terraces with low silting (Mytilineou et al. 2014;Bo et al. 2015), or on white coral mounds (Angeletti et al. 2014). Black corals are amongst the oldest known marine organisms, with a very slow growth and millennial longevities (Roark et al. 2009;Carreiro-Silva et al. 2013). In the Mediterranean, a validated age of a L. glaberrima colony caught off the Malta Island of approximately 650 years, with a radial growth rate of approximately 20 microns per year, was recently obtained by radiocarbon ( 14 C) dating analysis (Vitale et al. 2017).
Because of its smooth skeleton almost completely lacking in spines, L. glaberrima has been fished from ancient times and used for jewellery (Opresko 1998). Although the species was considered less valuable and more difficult to manipulate than red coral Corallium rubrum (Linnaeus, 1758), it was directly exploited by fisheries in some areas of the Mediterranean. From 1984 to 1987, 100-150 kg of black coral per year, presumably belonging to L. glaberrima, was caught by the St. Andrew's Cross towed on the bottom between 500 and 600 m south Gozo Island (Maltese Island) in the Strait of Sicily (Deidun et al. 2010). Despite the fact that today trade of L. glaberrima is illegal, after its listing among protected species by Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) Appendix II, the species is still impacted by deep-water fishing activities, being by-catch of bottom trawlers and longliners (Bo et al. 2014a;Deidun et al. 2014). Trawlers directly produce a reduction in the coral coverage on the swept bottoms, while long lines produce mainly indirect damage due to the abrasion by entangled gears on the colonies (Bo et al. 2014a). This abrasion facilitates bacterial infections and colonization of epibionts on the naked skeleton of corals, increasing coral mortality. On the other hand, lost fishing gear can accumulate on the sea bottom, covering wide areas and hindering coral recolonization.
Due to its main biological traits (sessile life and high longevity), typical of highly vulnerable and lowresilience organisms, the black coral L. glaberrima was recently proposed as a representative species for Vulnerable Marine Ecosystems (VMEs) in the Mediterranean (FAO 2016;Oceana 2016;GFCM 2017). Considering that the protection of VMEs has been a legal obligation for Regional Fishery Management Organizations since 2008 under the United Nations General Assembly (UNGA) Resolutions 59/25, 61/105 and 64/72, proposals to address the management of deep-sea fisheries and the protection of VMEs in the GFCM area were recently presented by GFCM (2017). These proposals included both the adoption of environmental and biological indicators to identify VME location, and the establishment of an encounter protocol to avoid fishing on these vulnerable habitats. Within this framework, knowledge on areas where L. glaberrima and other VME indicators are concentrated should be the basis of a spatial approach to fishery management for VME protection. Even if the increasing ROV investigations since the early 2000s have improved knowledge on L. glaberrima in the Mediterranean Basin, a synoptic distribution of these findings is still lacking. The aim of this paper is to provide a synoptic review of the available information on spatial distribution and main biological features of L. glaberrima in the Mediterranean. The paper is also intended to identify areas that can be considered hotspots of the species in the entire basin. This information is a prerequisite for any knowledgebased protection of habitats where L. glaberrima lives in particular, and for VMEs in general, through the establishment of new Fishery Restricted Areas (FRAs) in the Mediterranean Sea.

Materials and methods
In order to provide a synoptic view of L. glaberrima distribution in the Mediterranean, all the available geographical information with main biological features of this species in the Mediterranean was carefully checked. A systematic review based on common terms from the literature was carried out through electronic databases. Ten papers published in International Scientific Indexing (ISI) journals were found in the scientific database "Web of Knowledge" platform using the keywords "Leiopathes AND glaberrima AND Mediterranean". Other documents (non-ISI journals, conference abstracts and technical reports) were obtained by a wider search on "Google Scholar". The web search allowed us to identify 37 sources (excluding Esper (1788-1830) -Code 13, and our study -Code 22 in Table I) that comprised 31 peerreviewed papers, two technical reports and four conference abstracts.
A data set using all the scrutinized information such as locality, region of the Mediterranean, authorship, depth, abundance, and investigation gear/tool was created and is reported in Table I. Data on abundance were ranked on a three-level scale ("presence"less than 0.05 colonies per m 2 ; "colony"up to 0.2 colonies per m 2 ; "forest"more than 0.2 colonies per m 2 ) and are mapped in Figure 1. The geographical coordinates reported in the literature were standardized or, where lacking, extrapolated from locality and depth description.
Owing to the importance of other benthic invertebrates as indicators of VME (Oceana 2016;GFCM 2017), their co-occurrence with L. glaberrima findings is reviewed and listed in Table II, distinguishing the threatened taxa reported in the International Union for Conservation of Nature (IUCN) Red List Categories.
Data from the literature were enriched by two new findings collected by trawling and deposited in the Consiglio Nazionale delle Ricerche, Istituto per l'Ambiente Marino Costiero (CNR IAMC) collection of Mazara del Vallo (Sicily), mapped with a different symbol and discussed in detail below.

Mediterranean distribution and main associated fauna
The species L. glaberrima was described for the first time by Esper (1788Esper ( -1830 at the end of the 18th century, analyzing sample from the Gulf of Naples (southern Tyrrhenian Sea).
Since then, as listed in Table I and mapped in Figure 1, 37 records of the species (excluding those of Esper (1788Esper ( -1830, for the description of the species -Code 13, and of our study -Code 22) have been reported in the Mediterranean, especially since the early 2000s after the increase of ROV explorations.
The records available for the central basin of the Mediterranean are more numerous than for the western and eastern ones. In the central Mediterranean the "presence" of the species has been reported in the Corsica Sea, in the Strait of Sicily (Graham Shoal), in the northern Ionian Sea off Apulian coasts and in the eastern Ionian Sea (Cephalonia and Zakynthos Islands); and as a "colony" in the Ligurian Sea (S. Lucia Bank), in the Sardinian Channel (Skerki Bank), in the Tyrrhenian Sea (Vedove Shoal, Montecristo, Filicudi and Levanzo Islands, Cape San Vito), in the northern and eastern Ionian Sea (off Cape S. Maria di Leuca and Cephalonia Ridge) and in the southern Adriatic Sea (off Montenegro and Greece). It is worth noting the occurrence of pristine L. glaberrima "forest" in the Sardinian Sea (Carloforte Shoal), in the southern Tyrrhenian Sea (Marco Bank off Northern Sicily and off Pontine Islands -Palmarola and Zannone) and in the Strait of Sicily (Malta Graben).
In the western Mediterranean, only the "presence" of L. glaberrima was found, in the Alboran Sea (Cabliers and Chella Banks), in the Menorca and Majorca Channels (Balearic Islands) and in the Gulf of Lion (Bourcart and Cassidaigne Canyons).
Concerning the eastern Mediterranean, only the "presence" or "colony" of the species was reported, in the Aegean Sea (off Rhodes Island).
The benthic invertebrates living in the same habitats colonized by L. glaberrima and indicators of VME were found in 28 Mediterranean areas and are synthesized in Table II.

Discussion
A review of the literature showed that L. glaberrima is distributed in the Mediterranean from Cabliers Bank in the Alboran Sea (Western Mediterranean) to Rhodes Island in the south Aegean Sea (Eastern Mediterranean), with the highest level of sightings around the Italian coasts (Central Mediterranean). Although its presence on deeper bottoms could not be excluded in the Mediterranean, the depths at which the occurrence of the species was documented by scientific investigation ranged between 70 m (western Carloforte Island in the Sardinian Sea - Angiolillo et al. 2015) and 790 m (off Cape S. Maria di Leuca in the northern Ionian Sea - Figure 1. Distribution of Leiopathes glaberrima in the Mediterranean (circle: findings from the literature; star: findings from the present study). Abundance was ranked on a three-level scale: "presence"less than 0.05 colonies per m 2 ; "colony"up to 0.2 colonies per m 2 ; "forest"more than 0.2 colonies per m 2. Mastrototaro et al. 2010). However, D'Onghia et al. (2016, on the basis of interviews with fishermen, reported that L. glaberrima colonies or fragments have been caught from a minimum of 45 m by gillnet to a maximum of 1000 m by longlines. It may be noted that different sampling approaches can affect knowledge on coral beds' features. While the density of black coral colonies and associated fauna can be estimated by ROVs with good accuracy, longlining and trawling can be considered accurate mainly for information on the occurrence of epi-and endo-bionts. Considering that 25 out of 37 of the consulted papers are based on ROV investigations, we are confident that our review of information in the literature allows us to identify some areas where L. glaberrima is highly concentrated in the Mediterranean Sea. More detailed investigation is needed to produce accurate maps of the habitats occupied by L. glaberrima in these areas; nevertheless, hotspots of the species were reported off the Carloforte Island in the Sardinian Sea , on the Marco Bank and in the Pontine Islands (Palmarola and Zannone) in the Southern Tyrrhenian Sea (Bo et al. 2014a), and on the Malta Graben in the Strait of Sicily (Deidun et al. 2014) (see Table I).
It must also be emphasized that in the above areas between 70 and 90% of the benthic invertebrates associated to L. glaberrima are VME indicators, most of them being cnidarians. Furthermore, the role of forests of L. glaberrima as an essential habitat for completing the life cycle of commercial fish species was outlined by Cau et al. (2017). These authors reported a large number of egg capsules of spotted catsharks Scyliorhinus canicula (Linnaeus, 1758) anchored exclusively to colonies of L. glaberrima off the Carloforte Island, with an average of five capsules per colony and a maximum of 44 capsules in a 2-m-tall colony.
The protection of VMEs became a legal obligation for Regional Fishery Management Organizations, and the Food and Agriculture Organization of the United Nations (FAO) General Fishery Commission for the Mediterranean (GFCM) established four offshore FRAs to protect deep-water VMEs: the Eratosthenes Seamount, the cold hydrocarbon seeps of the Nile Delta, the white coral reefs off Capo Santa Maria di Leuca and the canyons in the Eastern Gulf of Lion. Comparing the existing FRAs with the documented spatial distribution of L. glaberrima, it is evident that, with the exception of the protected areas off Capo Santa Maria di Leuca and on the slope of the Gulf of Lion (see Table I), the black coral VMEs in the Mediterranean are not yet subject to specific protection measures. According to Auster et al. (2011), the biggest constraint in the process to protect VMEs is the uncertainty about the distribution and abundance of VME indicator species, and about the link between fishing efforts and significant adverse impacts. However, in the last decade the Vessel Monitoring System (VMS) and the Automatic Identification System (AIS) have become powerful tools for spatial-based management of fishery activities (Russo et al. 2011). Within this context, our review provides a first inventory of the distribution of L. glaberrima in the Mediterranean, suggesting that the presence of deep-water black corals may be more common in the area than assumed until now. Furthermore, data scrutinized in this paper can be used for modelling of habitat suitability of L. glaberrima or other antipatharians, and to contribute to an estimation of the potential occurrence of these species in the Mediterranean (Giusti et al. 2014;Etnoyer et al. 2017;Lauria et al. 2017).
In data-poor conditions, an integrated approach collating information from geo-surveys, ROV observations, trawl surveys, and commercial fisheries with habitat modelling could represent a powerful tool to identify the spatial distribution of VMEs and to promote effective conservation initiatives. This information could be increased by the adoption and implementation of a precautionary encounter protocol in bottom fisheries based on indicator species, including scientifically based thresholds for triggering a move-on rule (GFCM 2017).