**4. Discussion**

#### *4.1. Distribution of Eunicella verrucosa and Population Singularity*

Although recently the Sardinian coasts have been widely explored through a series of ROV surveys [40–44], no colonies of *Eunicella verrucosa* have been recorded. This evidence agrees with the recent map of the species distribution at the Mediterranean basin scale published by Chimienti [28], which includes data from original investigations, scientific literature and citizen observations validated by photographs.

The record of a persistent population settled in the Tavolara area for at least several decades, is, in this light, a peculiar feature of this zone. Its occurrence in the TPCCMPA was already observed in some previous investigations [29–33].

Here, the *E. verrucosa* colonies settle within sponge-dominated assemblages present on granitic outcrops under a high sedimentation rate in the Tavolara Channel, where the development of crustose coralline algae is limited. A similar assemblage including *E. verrucosa* together with *Axinella polypoides* and massive sponges was described on silted rocks of various lithology at 40–70 m depth in many sites along all Ligurian Sea [8,28].

Our observations indicate that the *E. verrucosa* population is composed of scattered colonies without formation of true forests; in fact, the recorded 100 colonies were grouped on 34 rocky outcrops reaching a total area of about 2 ha. Nevertheless, the size structure of the population seems equilibrated, with a symmetric size–frequency distribution and the modal class of the distribution, 30–40 cm, completely overlapped with that recorded by Chimienti [28] for the denser forests of Sanremo (Ligurian Sea). The estimated age structure reflects previous data obtained in the Marseille region. The data recorded for Sanremo population was younger (11–15 years), with a modal class in the range of 26–30 cm and a tail of old colonies reaching a maximal estimated age of about 71–75 years [28]. It is probable that sexual reproduction is only possible in a cluster of outcrops very close to the central of the channel. The maximal settling distance for this center was about 1.2 km for colonies recorded during this study and about 2 km for the colonies settled on the Klearchos wreck [30]. These data agree with the behavior of the lecithotrophic larvae of *E. verrucosa* showing a dispersion ability around the parent colonies <1 km [5,19,21,45].

The occurrence of several, well-developed colonies on the hull of the Klearchos wreck, sunk on 20 July 1979 [30], is a useful opportunity to validate the age estimation of the population. Although no reference scale is present in the available images, recorded in September 2011, the age of the wreck is in accordance with the modal class of the age distribution recorded during our survey.

The overall rarity of this species in Sardinia is difficult to argue, in the light of the wide occurrence of all the other shallow-water and mesophotic species of alcyonaceans [40–44].

The map of distribution proposed by Chimienti [28] suggests that the predominant Mediterranean water circulation explains the gradual colonization of the western Mediterranean Sea by *E. verrucosa* from the Atlantic Ocean. The presence of this coral in both the Balearic Sea and the Strait of Sicily may be explained by some common environmental and oceanographic features of these two areas. Both are characterized by an intense geostrophic circulation of water masses and a complex seafloor topography, that, due to the presence of islands and seamounts, generates mesoscale eddies and convergen<sup>t</sup> fronts [46–51]. In the Balearic Sea, the colonisation is driven by the ascending Atlantic Water (AW, surface water of Atlantic origin), which, bordering the western coast of Corse, where populations were recorded, see [28], enters the Ligurian Sea, reaching the Tuscany Archipelago. In this area, coral larvae can be spread by the Lyon Gyre [52].

On the other hand, from the Sicily Channel, the species has colonised the Tyrrhenian coast without going beyond the Gulf of Naples. Therefore, the species appears absent in the central Tyrrhenian Sea and the population of Tavolara MPA could be the unique description for the entire sub-basin. In this situation, it is plausible that the occurrence of this species at Tavolara could result from a stochastic event of settling of larvae presumably coming from the Tuscany Archipelago. Genetic studies on this isolated population might help to clarify its origin and connectivity with other coastal forests.

#### *4.2. Predators and Acrophilic Epibionts*

The study of the associated community provides some data about the specialized predators of *E. verrucosa*. The tritoniid nudibranch *Duvaucelia odhneri* lives its entire life cycle on the same host colony, exploiting seven different gorgonian species [53], including *E. verrucosa*, as also confirmed in this investigation with the discovery of individuals and eggs on the same colony.

The ovulid *Simnia spelta* shows a similar life strategy, being associated with at least four gorgonian species, *Eunicella cavolini*, *E. singularis*, *L. sarmentosa*, and, in our case, *E. verrucosa. S. spelta* feeds on the coenenchyme and polyps of the host and also lays ovarian capsules on the branches, causing necrosis of the underlying tissue [54].

During this study, we observed one colony of *E. verrucosa* hosting a group of the palaemonid shrimp *Balssia gasti*. The species has always been observed associated with octocorals, although the nature of the association is still to be elucidated. However, a predatory strategy was hypothesized due to the homocromic camouflage of this species in agreemen<sup>t</sup> with the color of the coenenchyme of the hosts [55–58]. In the Tavolara area, *B. gasti* was already observed on *Paramuricea clavata* and *E. cavolini* [59]. Finally, two specimens of the decapod *Periclimenes scriptus*, known as mucus-feeder of octocorals [56], were recorded. In the Tavolara area, this species was observed mainly associated with *P. clavata*, living on the granitic outcrops of the Tavolara Channel [60].

Regarding acrophilic species, the frequent occurrence of the basket star *Astrospartus mediterraneus* is remarkable. This species, generally recorded as colonizing deep habitats, [61] is becoming more and more abundant in relatively shallow waters in recent years [62]. *A. mediterraneus* is one of the few species that have changed its bathymetric distribution moving towards the surface. This is unusual; in fact, in relation to water temperature increasing, numerous shallow-water species changed their bathymetric distribution, reaching deeper levels [63].

#### *4.3. Epibiosis and Health Status of the Population*

Out of 100 colonies, 45 showed damages of various entities, from small portions of naked skeleton to completely dead colonies (Figures 5 and 6). The main stressors able to influence the coenenchyme integrity of structuring anthozoans are thermal stress and mechanical injuries, mainly due to fishing activity [34,35,64–66].

Diseases due to heating phenomena with consequent necrosis of coenenchyme are widely documented in the TPCCAMPA for *Paramuricea clavata* and *Eunicella cavolini*, but not at depths exceeding 35–40 m, where the process of necrosis starts from the apical portions of the colonies [67]. Based on this evidence, thermal stress has a negligible influence on the damages described in this study. In fact, the population of *E. verrucosa* mainly lives below the depth where these phenomena are usually documented. Moreover, the degenerative processes due to thermal anomalies are evident as a naked portion of the apical branches. This kind of damage was only sporadically recorded in our study. The direct observation of lost lines entangled in the colonies suggests, at least partially, an anthropic involvement. Physical contact with fishing gear scrapes the gorgonian coenenchyme, favoring the development of epibionts. Epibionts substantially modify the host–environment interactions (e.g., transference of energy or matter), eventually reducing their fitness [68]. Large masses of epibionts lead to a burdening of the colonies and greater mechanical stress, increasing their resistance to water movement [26,34,35,69–71].

The most common epibionts observed in this study were *Alcyonium coralloides*, bryozoans (*Turbicellepora avicularis*, *Adeonella calveti* and *Pentapora fascialis*) and the demosponge *Crella elegans* (Figures 4B and 5).

*A. coralloides* is one of the first colonizers of the skeletal portions deprived of coenenchyme and can subsequently continue its expansion to the detriment of the coral to fully occupy its skeleton. On the Tavolara specimens, *A. coralloides* always settled in the basal or central portions of the fan and never on the apical parts. It is generally accepted that the high occurrence of this epibiont may be correlated, in frequented sites, to anthropogenic damages [72], supporting its use as a bioindicator of stress in coralligenous assemblages [26,34,73,74]. Similar considerations are also true for erect bryozoans, such as *T. avicularis* [75].

Many Mediterranean localities endure impacts by anthropogenic pressure due to demersal fishing activities that pauperize three-dimensional benthic ecosystems, such as coral forests [10,26,35,40,69,76–81]. In the Medes Islands (Catalan Sea), between 10% and 33% of the colonies in unprotected populations were partially colonized by epibionts, most likely following tissue injury, whereas only from 4% to 10% of the populations in a marine protected area was affected [76], suggesting that fishing activities directly cause severe damage expressed as epibiosis coverage. Our data indicate an epibiosis at least four times higher for the Tavolara's *E. verrucosa* population.

The communities occurring on the granite outcrops have traditionally been considered of low quality due to the absence of the typical coralligenous features such as the thick coralline algal concretion [82]. This underestimation serves to classify the Tavolara Channel in the C-zone of MPA (Partial Reserve), allowing artisanal and recreational fishing activities, which, in turn, probably increased the pressure on the benthic communities. Recently, however, it was stated that the communities settled in this particular habitat are not impoverished facies of the coralligenous assemblage, but a peculiar community composed of erect sponge and habitat-forming anthozoans [33]. An adjustment of the managemen<sup>t</sup> guidelines of the MPA is required in light of the re-evaluation of this habitat.

**Author Contributions:** Conceptualization, G.B., M.C. and E.T.; methodology, M.C. and E.T.; validation, M.C., M.B. and F.E.; formal analysis, M.C.; investigation, M.C. and E.T.; resources, E.T.; data curation, E.T.; writing—original draft preparation, M.C., E.T., M.B., F.E. and G.B.; writing—review and editing, M.C., E.T., M.B., F.E. and G.B.; visualization, M.C. and E.T.; supervision, M.B. and G.B.; project administration, G.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data is available upon request form a corresponding auhor.

**Acknowledgments:** The authors would like to thank the Tavolara MPA managemen<sup>t</sup> for the permission to use the GIS environment and for the possibility to independently develop the underwater surveys. The authors would like to thank the "Slow dive" team for its support during the diving activities. Special thanks go to Florian Holon and Laurent Ballesta of ANDROMEDE Oceanologie, for providing the photographic material on the Klearchos wreck.

**Conflicts of Interest:** The authors declare no conflict of interest.
