Volcanic-associated ecosystems of the Mediterranean Sea: a systematic map and an interactive tool to support their conservation

Scoping and keyword string definition

Different search strings were tested to identify the most appropriate database for the literature data analysis. We started with a scoping stage based on the keywords “hydrothermal vents” AND “Mediterranean sea”. To help define the keywords, we used an adaptation of the Population, Intervention, Control, Outcome (PICO) framework (Collaboration for Environmental Evidence, 2022). Here we used PICo (Population, Interest, Context), identifying the volcanic-associated ecosystems as our population, the evidence existing in the literature as our Interest and the Mediterranean Sea as the Context.

The final literature search included three substrings connected using Boolean operators (AND and OR) and the wildcard “*” (Table S2):

‘(“hydrothermal*” OR “emission*” OR “volcan*” OR “plume*” OR “vent*” OR “seep*” OR “eruption*” OR “acidification” OR “carbon dioxide” OR “pH” OR “CO₂” OR “CCS”) AND (“Mediterranean”) AND (“sea” OR “ocean” OR “marine”)’.

The substrings used were broad enough to collect a large amount of literature however limiting the results in line with the objective of the research (James, Randall & Haddaway, 2016). The final search string was slightly modified depending on the database used (Table S2). Finally, we specifically searched for the scientific articles directly related to some of the projects focused on volcanic-associated ecosystems carried out totally or partially in the Mediterranean Sea: MedSeA (Mediterranean Sea Acidification in a changing climate, 2011–2014, http://medsea-project.eu/), EPOCA (European Project on Ocean Acidification, 2008–2012, https://cordis.europa.eu/project/id/211384), BIOACID I, BIOACID II and BIOACID III (Biological Impacts of Ocean Acidification, 2009–2012, 2012–2015, 2015–2017, https://www.bioacid.de), MIDAS (Managing Impacts of Deep-seA reSource exploitation, 2013–2016, https://www.eu-midas.net/), HERMES (Hotspot ecosystem research on the margins of European seas, 2005–2009, https://cordis.europa.eu/project/id/511234) and HERMIONE (Hotspot Ecosystem Research and Man’s Impact on European seas, 2009–2012, https://cordis.europa.eu/project/id/226354) (Table S2).

Database and searches

The Scopus database and the platform Web of Science were used to collect the scientific literature. These databases were chosen to ensure the reliability of the gathered information, since only indexed and peer-reviewed publications are allowed, disregarding all the non-indexed works. It should be noted that only English language literature was searched for and retained. Reviews were retained only when new results were presented, and the reported references were analyzed to include missing studies in the SM.

An additional search was performed on Google Scholar, and with a screening of the first 100 results (Haddaway et al., 2015) (Table S2).

Exported results

Search results from Scopus and Web of Science were exported in .csv format along with all the literature information including abstract and keywords. Literature search results from Google Scholar were manually added to a spreadsheet and all exported files were then loaded in the R environment (version 4.2.2; R Core Team, 2022) in RStudio (version 2022.12.0; RStudio Team, 2022) using the revtools package (Westgate, 2019).

Duplicate removal

As literature searches were performed on multiple online tools, some publications (from now on referred to as items) might be present multiple times. Therefore, the DOI (Digital Object Identifier) was used to identify and remove duplicates using the function find_duplicates in the revtools package and to create a database of unique studies.

Search results and screening

Overall, 10,310 items were identified from the online database searches: 5,472 and 4,838 from the Scopus and Web of Science searches, respectively. In addition, the searches of scientific articles related to projects (MesSeA, EPOCA, BIOACID I, BIOACID II and BIOACID III, MIDAS, HERMES and HERMIONE) returned 858 items. Across the combined searches, a duplication rate of 36.4% was estimated and the number of remaining articles was 7,100. The title and abstract screenings excluded 5,691 articles. All the remaining literature items were retrieved (using the access provided by all the co-authors’ institutions) and the full-text screening was then carried out. Finally, 433 literature items were coded in the SM database with a total of 992 observations (last update 27/05/2022; Fig. S1). The complete list of literature items included in the database is provided in the Supplemental Material (Table S3).

Dates, study types and journals analysis

The earliest article in the database refers to a work published in 1973, followed by an increasing trend in the number of publications showing a peak in 2014 (Fig. 1). The selected studies included four document types, with a total of 419 articles (96.8% of total studies), three conference papers (0.7%), eight proceedings papers (1.9%) and three reviews (0.7%) (Fig. 1). The studies were published in more than 100 different journals with only five journals that published more than 20% of the total: Marine Geology (8.6%), Chemical Geology (4.4%), Marine Environmental Research (3.0%), Science of the Total Environment (3.0%) and Deep-Sea Research Part I: Oceanographic Research Papers (2.8%) (Fig. 2).

(Bio)geography and population analysis

There are twenty-one regions in the Mediterranean Sea that include hydrothermal vents, cold seeps, mud volcanoes, pockmarks and seamounts with 156 unique sites between 1 and 3,800 m depth (Fig. 3). A list of all the sites is reported in Table S4.

Almost 50% of the total number of observations was located in the Italian Maritime Region (n = 486), followed by Greece (n = 221) and Turkey (n = 121). Less than 30% of the observations (n = 279) fall within some kind of protection measures by European, National or Regional regulation. The highest number of scientific observations was registered in the “Aeolian Archipelago, terrestrial and marine areas”, a region identified as a Special Protection Area (SPA) under the EU Birds Directive (2009/147/EC, European Union, 2009) (n = 166) (Fig. 3 and Table 2).

The highest fraction of observations identified falls into Hydrothermal Vents (47.9%), followed by Mud Volcanoes (30.7%) and Cold Seeps (5.8%) (Fig. 4A; Table S5). A slightly higher number of observations was reported in shallow water (<200 m depth) than in deep sea (>200 m depth; Fig. 4B).

Three regions included more than 50% of the total number of observations with the highest number recorded in the Aeolian Arc (27.9%), followed by the Gulf of Naples (14.7%) and the Aegean Arc (11.7%) (Fig. 4C).

Generally, more attention was given to the environmental characteristics of the habitat (57.8%) than the biological aspects (42.2%), but a higher number of biological observations was recorded in shallow water than in deep sea (n = 284 and n = 133, respectively, Fig. 4D).

The highest number of environmental observations involved the analysis of sediment (56.9%), followed by water (14.0%) and gas (2.3%), with the rest of observations involving a combination of the three abiotic matrices (Fig. 4E). In the Aeolian Arc a higher number of observations involved the analysis of environmental characteristics (54.3%) than the biological components of the systems (45.6%).

Ten main biological targets were identified in the literature analyzed: algae, bacteria, epibenthos, epiphytes, fish, macrobenthos, meiofauna, plankton, seagrass and virus. The greatest number of observations was focused on macrobenthos (39.1%), followed by bacteria (26.6%) and algae (10.3%). The highest number of observations within the analyzed categories was related to shallow-water areas, with more than twice the number of observations for deep-sea areas (Fig. 4F). The highest number of biological observations involved the survival response (83%), with the other responses covering less than 20% of the total observations (Fig. 4G).

Keywords analysis

More than 900 unique keywords in the database were identified. The most frequent keywords were “ocean acidification” (n = 53), “mud volcano” (n = 35) and “mediterranean sea” (n = 34), followed by “cold seep” (n = 29) and “hydrothermal vent” (n = 16), while all the other keywords were less mentioned in the whole database (Fig. 5A).

We also examined the results of the co-occurrences of author’s keywords using a network visualization graph to identify patterns of relationships between keywords. The keywords “mediterranean sea” and “ocean acidification” mostly frequently occur together creating a cluster at the centre of the representation. In contrast, keywords related to “authigenic carbonates” in the “Sea of Marmara”, “metalliferous sediments” in the “Aeolian Arc” or “microbial biofilms” remained at the margin of the network (Fig. 5B).

A tool for managers: the MH-shiny app and its interactive map

In the context of the FAIR principles (Findable, Accessible, Interoperable and Re-usable data) of the European Commission “Open Data Directive” (European Union, 2019), the MH-shiny has been developed during this study (cfr. link: https://costavale.shinyapps.io/MH-shiny/) as a shiny-based application freely accessible online. The complete R code and the data are freely available on a GitHub repository (https://github.com/costavale/MH-shiny/; DOI 10.5281/zenodo.7537047). The shiny app works on both local and online versions on macOS, Windows, and Linux operative systems.

The interface of MH-shiny consists of three main sections (Fig. 6; Figs. S2–S4). The first section is a user-friendly “Interactive Map” where the user can select the country, region, site or site type directly on the map (Fig. S2). The selection will automatically connect to the second section, the “Data Explorer” which shows the data as a list of the literature items and a graphical representation of the variables (chosen by the user), which can be directly downloaded as a .csv file or as a .png image, respectively (Fig. S3). The third section “Keywords Analysis” provides a word-cloud analysis of keywords (author or index keywords) as a visual representation of the most frequently used, a graph of the number of occurrences of the keywords and a network graph of co-occurrence keywords (Fig. S4).

The Aeolian Archipelago case study

Despite 30% of the volcanic-associated ecosystems are enclosed within ABMTs for conservation, not all of them are protected. To explain this discrepancy, we use the case study of the Aeolian Archipelago, a volcanic arc located in the southern Tyrrhenian Sea. This region is characterized by an exceptional marine biodiversity due to the elevated number of different habitat types and organisms present in the area (Consoli et al., 2021). Marine hydrothermalism is a diffuse phenomenon in the whole area and the entire volcanic arc is the most scientifically explored in the Mediterranean Sea because of the heterogeneity of its hydrothermal structures (Dekov & Savelli, 2004; Lupton et al., 2011; Esposito et al., 2018; Rizzo et al., 2022). The archipelago has an extension of 22 km² and consists of seven main islands (Alicudi, Filicudi, Salina, Lipari, Vulcano, Panarea, Stromboli) with associated islets (Basiluzzo, Dattilo, Bottaro, Lisca Bianca, Strombolicchio), and several seamounts (Beccaluva et al., 1985; Gamberi & Marani, 1997; Lupton et al., 2011). The area holds the status of Special Conservation Area (ITA030044 ZPS-RETE NATURA 2000 “Arcipelago delle Eolie—area marina e Terrestre”) based on the EU Birds Directive (2009/147/EC) and is therefore identified as Natura 2000 site and has been listed as UNESCO World Heritage Site (Angelini, 2008). The establishment of a national Marine Protected Area that covers the entire archipelago, already planned by the Italian law 979/82, has been underway for decades. In 2014, the Sicily region in collaboration with UNESCO delivered a proposal of a management plan to overcome the mismanagement of the natural, geophysical and archaeological heritage of the archipelago, supporting the establishment of the Marine Protected Area (MPA) (Angelini, 2008).

We identified fourteen different important habitat types characterizing the volcanic-associated ecosystems in the area, along with several protected species of algae and invertebrates living in association with the hydrothermal vents (UNEP, 1973; Council of Europe, 1979) (Table S6). Among these habitats, several are listed as priority habitats in the Annex I of the Habitats Directive (92/43/EEC) and in the Protocol Concerning Specially Protected Areas and Biological Diversity in the Mediterranean of the Barcelona Convention (SPA/BD Protocol, UNEP/MAP, 1995; Consoli et al., 2021) (Table S6). According to the SPA/BD Protocol, the conservation of these habitats is mandatory (SPA/BD protocol of the Barcelona Convention, UNEP/MAP, 1995).

Despite the efforts to improve the protection of the marine environment surrounding the Aeolian Archipelago, the entire area is still subjected to significant anthropogenic pressures deriving especially from touristic and artisanal fishing activities on which the archipelago’s economy strongly relies. Derelict fishing gears and general waste from land pollution and touristic activities (e.g., plastic bottles, metals, ceramics, glass) have been found associated with hydrothermal sites around the islands where entanglement and ghost fishing have been documented also, providing the unquestionable evidence of the anthropogenic impacts even in the deep sea (Consoli et al., 2021). Moreover, years of unmanaged scientific research that left instruments and/or used destructive sampling methodologies could represent a source of additional environmental disturbance in the area (Dando & Juniper, 2001).

Several investigation have been conducted in the archipelago providing data on the geochemical (Italiano & Nuccio, 1991; Sedwick & Stuben, 1996; Capaccioni, Tassi & Vaselli, 2001; Price & Pichler, 2005; Italiano, 2009; Boatta et al., 2013; Price et al., 2015) and biological (Calosi et al., 2013; Apostolaki et al., 2014; Johnson et al., 2015; Harvey et al., 2016; Vizzini et al., 2017; Mirasole et al., 2020; Noè et al., 2020) setup of many hydrothermal vent sites of this area. Here, marine hydrothermalism is characterized by sporadic and unpredictable underwater phenomena such as the 2002 massive underwater explosion near Panarea Island (Esposito, Giordano & Anzidei, 2006). Such natural hazard phenomena can severely impact the marine environment through the release of great amounts of heavy metals and trace elements in the surrounding habitats, potentially causing bioaccumulation in the local fishing population (Andaloro et al., 2012). This is a public safety concern that needs to be managed appropriately in order to avoid damage to the local population, fishers, divers, or tourists (Esposito, Giordano & Anzidei, 2006; Aliani et al., 2010; Vizzini et al., 2013; Mishra, Santos & Hall-Spencer, 2020; Voltattorni et al., 2006).

Considering the detrimental impacts that the above-described commercial and scientific activities can have on morphological, geochemical, and biological aspects of the ecosystems along with the occurrence of protected species and priority habitats, the gas hazard and the related environmental contamination risk, the setting and implementation of management measures in line with the current conservation policies in this region is urgently needed to protect and manage these peculiar ecosystems.

The creation of a MPA (according to Italian Law 979/82, Art. 31) where all activities are strictly regulated could be a solution, with no-take/no-access zones regimes (i.e., integral reserve) at least for the most sensitive hydrothermal sites (Esposito et al., 2018; Aiuppa et al., 2021; Consoli et al., 2021). The designation of Site of Community Importance (SIC) or Special Area of Conservation (SAC) would also be legitimated by the presence of critical habitat types or species as respectively listed in Annex I and II of the European Habitats Directive (http://ec.europa.eu/environment/nature/legislation/habitatsdirective/index_en.htm#enlargement). The MPA would be the best bet to manage multiple uses of the archipelago with conservation as a priority objective, followed by the allocation of areas for regulated scientific research and monitoring, in light of the strong need to fill knowledge gaps on the present volcanic ecosystems. Implementing up-to-date protection measures in the Aeolian Archipelago would also increase people’s awareness of the importance of preserving marine hydrothermal vents based on their high naturalistic importance, thereby encouraging the development of sustainable activities for fishing and tourism in the whole area. An ecosystem-based conservation strategy is required to identify the priority criteria for the protection of these volcanic-associated ecosystems (Fanelli et al., 2021).