Spring composition of the macroalgal vegetation of a small offshore island in the north-western Mediterranean (Gallinara Island, Ligurian Sea)

Gallinara Island, a small island located 1.5 km off the shore of Liguria (Italy, north-western Mediterranean Sea) was included in a list of proposed Marine Protected Areas (MPA) in the early 90s. Since then, its benthic assemblages have been studied in detail and the main macrophytic communities have been mapped. A detailed assessment of its benthic macroalgal flora, however, has never been made. Gallinara was visited in the course of 5 consecutive years and its macroalgal flora was studied based on collections made by snorkelling and SCUBA diving. Overall, 141 macroalgal taxa were collected and identified (23 Chlorophyta, 94 Rhodophyta, 24 Ochrophyta); 91 of them represent new records for the island. One of the most notable new records is the non-indigenous red alga Womersleyella setacea, previously unreported from the island and widely distributed, particularly on the south-eastern shore. Observations made in the course of the surveys confirm the rarefaction of some large-sized brown algae (particularly Sargassum vulgare) but indicate also that others previously reported as rare (Cystoseira compressa, Dictyopteris polypodioides) are still common on the island.


Introduction
Littoral systems around small islands often show peculiar characteristics, both biotic and abiotic, if compared with those of mainland coasts (Piazzi et al. 2004;Rindi and Guiry 2004). Differences in environmental factors such as wave exposure, currents, water transparency, sedimentation rate and interaction with pelagic systems are considered the main causes of these peculiarities (Piazzi et al. 2004). Such differences are usually more evident for islands located at greater distances from the mainland. In the last decades, anthropogenic impacts affecting many littoral habitats along continental shores have further contributed to increase these differences. This is a common situation in the north-western Mediterranean Sea, where the limited urbanization typical of many offshore islands has allowed the maintenance of a comparatively high environmental status, in contrast to the highly impacted continental shores in the region.
Gallinara is a small island located in the Ligurian Sea, approximately 1.5 km from the western shore of Liguria, in the Italian mainland. The island is not open to public access and its terrestrial biota, consisting mainly of Mediterranean shrublands, are generally well conserved, with the presence of rare and endemic species listed in Habitat Directive (e.g Campanula sabatia De Notaris 1844). Moreover, the island hosts one the main colonies of the seagull Larus michahellis Naumann 1840, of the Ligurian region. For this reason, a Natural Regional Reserve was established for the terrestrial environment in 1989 (Bianchi et al. 2018). The marine benthic communities of Gallinara were not studied in detail until the early 90s of last century (Balduzzi et al. 1992), when the island was included in a list of proposed Marine Protected Areas (MPA) in the framework of the Italian National Framework Law 394/91 on Protected Areas. To date a Marine Protected Area has not yet been established, but the island is included in a Special Area of Conservation (SAC "Fondali Santa Croce -Gallinara -Capo Lena", Natura 2000 code IT1324974). Despite of this, there is evidence that the rocky bottoms surrounding the island underwent major alterations due to increasing human impacts in the last 25 years, mainly in the form of anchoring and fishing, both artisanal and recreational (Bianchi et al. 2018).
The benthic macrophyte communities of Gallinara were characterized and mapped in several studies (Balduzzi et al. 1992;Tunesi et al. 2006Tunesi et al. , 2008Bianchi et al. 2018) that provided a background of knowledge for the establishment of the MPA. Macrophytes (seagrasses and benthic macroalgae) are considered useful descriptors of environmental characteristics of coastal habitats Balata et al. 2011). Attached algae, because of their sedentary condition, integrate the effects of long-term exposure to nutrients or other pollutants resulting in a decrease or even disappearance of the most sensitive species and its replacement by highly resistant, opportunistic species ). Therefore, these organisms are widely used for environmental assessments of coastal waters (Orlando-Bonaca et al. 2008;Orfanidis et al. 2011;Cecchi et al. 2014;D'Archino and Piazzi 2021) and the seaweed communities occurring in an area are an important aspect to take in account when the establishment of an MPA is considered.
Previous investigations have reported 71 species of macroalgae from Gallinara, but a detailed census of the macroalgal flora was not among their aims. In general, detailed recent studies focusing on the macroalgal diversity of the region where Gallinara is located (Liguria) are essentially lacking. We visited the island each year in a time span of 5 years (2013)(2014)(2015)(2016)(2017) and made targeted collections and observations on the benthic macroalgae. This improved substantially the knowledge of the macroalgal vegetation, providing new valuable floristic information. Observations of the seaweed communities also allowed to compare the present distribution of several species with information available in the literature.

Study site
Gallinara Island has a triangular outline and its perimeter measures approximately 1.4 km (Fig. 1). The coastline is entirely rocky and consists mostly of steep cliffs. With the only partial exception of the north-eastern shore where a small harbour is located, the island is characterized by wave-exposed conditions, due to exposure to southerly and south-easterly winds (that are the prevailing in the area). As in most of the Mediterranean Sea, the tidal range is limited and does not exceed 0.3 m (Bianchi et al. 2018). From the lithological point of view the island consists entirely of stratified quartzite (Bavestrello et al. 2000).

Methods
Gallinara Island was visited in June in 5 consecutive years (9-14 June 2013; 9-14 June 2014; 7-12 June 2015; 5-10 June 2016; 5-10 June 2017). Collections of benthic macroalgae were made from the mediolittoral, infralittoral and circalittoral zones, at depths ranging from the surface to approximately 45 m. Collections from the mediolittoral and shallow infralittoral (-1.5/2 m) zones were made by snorkeling along the whole shoreline of the island. Collections from the infralittoral and circalittoral zones were made from two sites, Punta Falconara (southern point of the island) and Punta Sciusciaù (eastern point of the island) by SCUBA diving. Sampling was based on targeted collections, made with the aim of collecting as many species as possible.
In addition, a visual census of the seaweed communities of the mediolittoral and shallow infralittoral zone (approximately 5 m depth) was carried out in June 2016. The whole perimeter of the island was checked by visual inspection and the species present were noted. For convenience, we present and discuss the results subdividing the island in 30 sectors of the four shores: north-eastern shore (NE, sectors 1-7), south-eastern shore (SE,, western shore (W, sectors 20-29) and harbour (H, sector 30).
Algal samples were removed from the substratum using a sharp knife or hammer and chisel (except for some samples from the infralittoral zone, which were collected using a vacuum pump). The material was placed in sealed plastic bags and examined a few hours after collection using a binocular stereoscope and a light microscope. Part of the material was immediately identified, dried, and pressed for preparation of herbarium specimens. Some specimens that could not be readily identified were stored in a solution of 4% formalin in seawater and identified upon return to the laboratory at the Polytechnic University of Marche. Identifications were based on morphological characters and were obtained using descriptions and illustrations mainly in Coppejans (1983), Cormaci et al. (2012Cormaci et al. ( , 2014Cormaci et al. ( , 2017Cormaci et al. ( , 2020Cormaci et al. ( , 2021 and Rodriguez-Prieto et al. (2013). Voucher specimens were deposited in the herbarium JIWUKORI of the Universidad Nacional de Colombia, Bogotá (JIW) and in the phycological section of the Herbarium Anconitanum, Università Politecnica delle Marche, Ancona (ANC). Algal nomenclature and taxonomy follow AlgaeBase (Guiry and Guiry 2022).

Results
Overall, 141 macroalgal taxa were collected and identified in the study (23 Chlorophyta, 94 Rhodophyta, 24 Ochrophyta) ( Table 1); 91 of these represent new records for the island (Table S1). In terms of distribution, 11 taxa were recorded from the harbour, 65 from the NE shore, 126 from the SE shore and 88 from the W shore (Table 1).   Dasya ocellata (Grateloup) Harvey + + + Not abundant, but frequently observed in the photophilous turfs mixed with other filamentous algae. Dasya rigidula (Kützing) Ardissone + + + Not abundant, but frequently observed in the photophilous turfs mixed with other filamentous algae.

Phylum and taxon Sector Distributional notes H NE SE W
Wrangelia penicillata (C. Agardh) C. Agardh + + + Lower mesolittoral and upper infralittoral; easily recognized by the whitish hairs present on the branches. Frequent, it is present mainly on the SE shore in sectors 10-14 and 16-17; observed also in sectors 22-24.
The composition of the algal assemblages of the infralittoral fringe (0.4/0.5 m) varied among different sectors of the shoreline, mainly in relation to morphology and inclination of the substratum. Well-developed populations of the fucalean brown algae Ericaria amentacea (C. Agardh) Molinari & Guiry and Cystoseira compressa (Esper) Gerloff & Nizamuddin occurred in some parts of the southern portion of the W shore and on the SE shore (where they occasionally formed stretches of continuous belt), whereas these species were scattered on the rest of the island. In parts of shoreline where these seaweeds were absent, the assemblages consisted of highly diverse turfs formed by articulated corallines ( (Table 1). Algal turfs were also common in shaded habitats, particularly the steep cliffs of the SE shore near Punta Falconara (sectors 14-18). In these habitats, however, their composition was different: the most common species were Acrosorium ciliolatum (Harvey) Kylin), Aphanocladia stichidiosa (  abundant (Fig. 2B) and thin articulated corallines (Jania spp.) were among the most common epiphytes. A peculiarity discovered in the shallow infralittoral zone was a population of the siphonous green alga Pedobesia simplex (Meneghini ex Kützing) M.J. Wynne & Leliaert formed by unattached specimens freely lying on sand in a sheltered inlet of the harbour (44°1'36.31"N, 8°13'29.56"E), at 0.5 m.
Deeper infralittoral and circalittoral zones were more heterogeneous. The parts of sea bottom towards the mainland (W-SW oriented) are characterised by a gentle slope and a patchy meadow of the seagrass Posidonia oceanica on sandy and coarse sediment, while toward the open sea (SE oriented) the island has steep rocky cliffs and landslide masses ending with detrital and muddy sediments at around 35 m depth, where a peculiar population of the boring sponge Cliona viridis (Schmidt, 1862) has been described in its gamma (massive) stage (Cerrano et al. 2007). The underwater seaward geomorphological profile at Punta Sciusciaù is characterized by vertical cliffs from the surface to 6 m depth, continuing with a plateau 20-30 m wide, with scattered rocks and pebbles with a very slight inclination down to 12 m depth, then from 12 to 24 m depth a second vertical cliff with some small terraces, followed by a gentler rocky slope from 24 m depth to 30 m depth, and then by scattered rocky blocks with coralligenous accretions. The underwater seaward profile at Punta Falconara is characterised by cliffs and landslides of boulders mixed to coarse sediments.

Discussion
With 91 new records, our results increase substantially the number of macroalgal taxa for which the occurrence on Gallinara is documented. Combining our results with information from relevant literature (Balduzzi et al. 1992;Mangialajo et al. 2004;Tunesi et al. 2006;Bianchi et al. 2018), the macroalgal flora of the island as currently known appears to consist of 162 taxa (28 Chlorophyta, 105 Rhodophyta, 29 Ochrophyta; Table S1). The number of taxa that we recorded is substantially higher than the estimate of 71 based on previous literature. Such increase is due mainly to the identification of numerous filamentous species that were not investigated in previous studies (which were based on visual census by SCUBA diving and ROV observations, methods inadequate to identify small-sized seaweeds). We believe, however, that the current estimate of 162 taxa is incomplete and probably not yet close to the actual total. Our investigations were carried out in early June, and therefore covered a limited span of the annual cycle. In the Mediterranean area early June is late spring; at this time of the year Mediterranean seaweed communities are highly species-diverse, with most species occurring in their fully developed habit. So, our sampling probably provided a good general representation of the macroalgal flora of the island. Nevertheless, it is obvious that our collections did not include seasonal species with temporal occurrence restricted to other periods (e.g., foliose red algae of the genus Pyropia, which in the Mediterranean occur in the mediolittoral zone in winter and early spring). In addition, our sampling could not cover the totality of benthic habitats around the island. In particular, we did not sample the meadows of Posidonia oceanica located in the harbour and near the northernmost part of the western shore. Posidonia oceanica meadows are three-dimensional environments characterized by a high associated biological diversity, with distinctive assemblages occurring on the leaves and on the rhizomes (Piazzi et al. , 2015. So, almost certainly we missed some species known only or mainly as epiphytes of the leaves of this seagrass (e.g., the brown algae Cladosiphon cylindricus (Sauvageau) Kylin, Giraudia sphacelarioides Derbès & Solier, Myriactula gracilis van der Ben, Myriactula rivulariae (Suhr ex Areschoug) Feldmann and Myrionema orbiculare J. Agardh). We also remark that our identifications were based entirely on morphological grounds. It is now well known that this approach is insufficient to unravel the full diversity of macroalgal floras and should be integrated with DNA sequence data (Bartolo et al. 2020;Rindi et al. 2020a). This is especially true for small seaweeds with simple morphology, e.g. foliose or filamentous, which abound in the Mediterranean macroalgal flora. Cases of cryptic diversity are frequent in these algae and have been documented in the recent past (e.g., Wolf et al. 2018Wolf et al. , 2019Díaz-Tapia et al. 2020;Serio et al. 2020). In recent years we produced DNA sequence data from samples collected from Gallinara, as part of separate studies focusing on some species (Pezzolesi et al. 2017(Pezzolesi et al. , 2019Caragnano et al. 2020;Rindi et al. 2020b). One of these (Caragnano et al. 2020) unravelled a striking case of cryptic diversity: we described the new species Lithophyllum pseudoracemus Caragnano, Rodondi & Rindi based on samples collected from a rhodolith bed at -40 m in the area of Punta Falconara. This cryptic species was discriminated from Lithophyllum racemus (Lamarck) Foslie on molecular basis; probably, records of Lithophyllum racemus from Gallinara reported by Bianchi et al. (2018) should be referred to it. For the purpose of the present study, unfortunately, we could not generate molecular data.
Despite of these limitations, our results recovered a relatively species-diverse macroalgal flora that can be considered typical of the north-western Mediterranean. The flora of Gallinara has substantial similarities with those of north-western Mediterranean islands and coastal areas that have been floristically well investigated. Of the 162 taxa reported for Gallinara, about 110 are common to Gorgona Island (Piazzi et al. 2004), about 100 to Giannutri Island  and 105-110 to the area of Calafuria, near Livorno (Rindi et al. 2002;). However, the actual number of taxa common with these locations is probably higher. In our study we could not identify at species level several algae, due to the small size of the specimens or the lack of critical diagnostic features (e.g., Bryopsis cf. duplex De Notaris, Bryopsis sp., Cladophora cf. dalmatica Kützing, Ulva cf. flexuosa Wulfen, Ulva cf. laetevirens Areschoug, cf. Acrosymphyton purpuriferum, cf. Callithamnion corymbosum (J.E. Smith) Lyngbye, Chondria cf. mairei Feldmann-Mazoyer, Gelidium cf. crinale, Liagora cf. viscida (Forsskål) C. Agardh) (for this reason, in the floristic comparison with the abovementioned localities, we report approximate numbers). It is likely, however, that these belong to species widespread in the north-western Mediterranean.
The spatial coverage of our study does not fully correspond to previous work performed on Gallinara (Balduzzi et al. 1992;Mangialajo et al. 2004;Tunesi et al. 2006Tunesi et al. , 2008Bianchi et al. 2018). Despite of this, our results allow some comparisons highlighting the recent changes that took place in the macroalgal communities. Our results confirm the rarefaction or loss of some large-sized brown algae, which seems to agree with the scenario of reduction of three-dimensional structure described by Bianchi et al. (2018). These authors reported a decrease of erect and canopy-forming species in parallel with an increase in filamentous algae, which caused homogenization of the epibenthic communities and destroyed the three-dimensional habitat structure that characterized Gallinara in the early 1990s. In particular, we did not record the brown seaweed Sargassum vulgare, either in our collections or visually, in any part of the island except the upper littoral rockpool of the SE shore. This macrophyte was common around Gallinara until 2009, when it used to form a conspicuous community in the infralittoral zone, best developed around 5 m depth (Bianchi et al. 2018). Its current rarity matches the decline described for the shores of Mediterranean France, for which habitat loss and pollution have been invoked as main causes (Thibaut et al. 2016). Similarly, we did not record two other fucalean brown algae previously reported from Gallinara, Ericaria zosteroides (C. Agardh) Molinari & Guiry (reported as common in the circalittoral of the SE shore by Balduzzi et al. 1992, as Cystoseira zosteroides; and still forming locally dense patches in the same area in July 2005, Mangialajo, personal communication) and Gongolaria montagnei (J. Agardh) Kuntze (reported from the W shore by Balduzzi et al. 1992, as Cystoseira spinosa). However, we did not investigate the deep infralittoral habitats occupied by these algae as carefully as the shallow infralittoral zone, so we may have missed them. Differently from Bianchi et al. (2018) we found that two other large brown seaweeds, Cystoseira compressa and Dictyopteris polypodioides, are still generally common on the island, although with distribution concentrated in certain habitats and certain parts of the shoreline (Table 1). These are two widespread Mediterranean phaeophytes that thrive in a wide range of environmental conditions and may also occur at sites moderately affected by anthropogenic disturbances (Mangialajo et al. 2008;Mancuso et al. 2018;Orfanidis et al. 2021). We also confirm the presence of Gongolaria elegans in the rockpool on the SE shore, where this species was previously recorded by Mangialajo (2007).
However, the most remarkable difference is the widespread occurrence of the filamentous red alga Womersleyella setacea (Hollenberg) R.E. Norris, which was not reported in previous studies (although we suspect that the filamentous algae noted by Bianchi et al. (2018) might be mainly this species). In the 5 years of our investigation, Womersleyella setacea was widely distributed in the lower infralittoral and circalittoral zones, where it formed dense filamentous turfs at depths ranging between 10 and 30 m, both at Punta Sciusciaù ad Punta Falconara. The widespread presence of this alga in the north-western Mediterranean was first noted by Verlaque (1989); this author remarked its abundance in the Rade de Giens (Var, France; a bay located about 150 km from Gallinara). Womersleyella setacea is considered an introduction in the Mediterranean; in recent decades, dense turfs formed mostly by this species have become a feature of the seascape of several north-western Mediterranean locations (Piazzi and Cinelli 2001;Piazzi and Balata 2009;Cebrian and Rodriguez-Prieto 2012). Microscopic observation of samples from Gallinara did not reveal any reproductive structures, supporting the idea that in the Mediterranean this alga reproduces only vegetatively (as suggested by previous studies, e.g. Airoldi et al. 1995;Athanasiadis 1997;Rindi et al. 1999;Cebrian and Rodriguez-Prieto 2012). The creeping axes of this species are capable to spread quickly over the substrata occupied, covering large portions of bottom in which thick turfs develop and recruitment of seaweeds reproducing sexually or by spores is hampered (Airoldi 1998;Piazzi and Balata 2009). Ballesteros et al. (2009) suggested the development of Womersleyella setacea as a likely cause of the decline of Ericaria zosteroides (as Cystoseira zosteroides) in the Scandola Marine Reserve (Corsica). This scenario appears compatible with the situation observed in Gallinara. A quantitative assessment of the distribution of this species on the SE shore of Gallinara will be presented in a separate manuscript (Torsani et al., in preparation).
We also recorded two other non-indigenous species, Caulerpa cylindracea and Acrothamnion preissi (Sonder) Wollaston, known for the invasive traits that they have shown in the north-western Mediterranean. Acrothamnion preissi was found in very small amounts, but it might be more common than we were able to detect; this filamentous species is a common epiphyte on the rizomes of Posidonia oceanica . Caulerpa cylindracea spread aggressively in the Ligurian Sea in the last 20 years and in 2015 was reported to be still increasing its range and habitat occupancy (Montefalcone et al. 2015;Cerrano et al. 2017). Its presence on Gallinara has been known since 2005 (Tunesi et al. 2006) and Bianchi et al. (2018) reported an increase in abundance of the species in the last decade. In the course of our surveys we observed this species in the in the rockpool of the SE shore and the infralittoral zone of the SE shore where, however, it did not appear to be abundant. Future surveys will be necessary to assess whether this is indicative of a regression of Caulerpa cylindracea in the area or not.

Conclusions
Our results confirm that some large-sized seaweeds considered important as habitat formers have become rare or disappeared from Gallinara. Despite of this, our investigations revealed a macroalgal flora that is substantially more diverse than appreciated so far, particularly in the shallow infralittoral zone. The number of species currently documented (which, as remarked above, is certainly incomplete) is considerable for an island with a 1.4 km-long coastline. The loss or rarefaction of a few iconic species does not seem to us sufficient to conclude that the macroalgal vegetation of Gallinara must be considered impoverished. In particular, the large number of filamentous, turfforming species, discovered should not be regarded as a reason to consider it a flora of low significance. There is a tendency to consider algal turfs as communities of lower value compared to canopy-forming species, and to treat all turfs as functionally equivalent units, regardless of their composition. In particular, most ecophysiological studies of algal turfs treat them at assemblage level and their species composition is often neglected (Provera et al. 2021). In fact, algal turfs may be very species-diverse and different turf-forming species, although morphologically similar, may respond differently to environmental change (Provera et al. 2021). So, even this component of benthic assemblages is important to study in detail in its all aspects (which has not been done for Gallinara so far).
Based on our results, we suggest that Gallinara is an interesting representative site that deserves further detailed investigation in relation to its benthic communities. In a more general perspective, considering the peculiarity of the island biodiversity, we also argue that plans for the creation of an MPA should be revitalized, in agreement with the urgent need to protect and conserve at least 30% of the Earth's lands and waters by 2030 as expected by the EU's biodiversity strategy.