Chondrichthyes are one of the oldest and most ecologically diverse vertebrate lineages. They appeared about 420 million years ago occupying the upper levels of aquatic food webs (Compagno 1990, Kriwet et al. 2008). Today, they have a conspicuous number of predators playing an important role within the marine ecosystem (Stevens et al. 2000, Ferretti et al. 2010, Heithaus et al. 2012), so if these species disappear, the balance within the aquatic food webs would be compromised. It is well known that overfishing and habitat degradation have altered marine animal populations (Polidoro et al. 2012), especially those of sharks and rays (Ferretti et al. 2010, Dulvy et al. 2014). Cartilaginous fish are intrinsically sensitive to high fishing mortality because of their K-selected strategy (e.g. slow growth, late attainment of sexual maturity, long life spans and low fecundity; Stevens et al. 2000). They are frequently captured incidentally, but are often retained as valuable by-catch of fisheries that focus on more productive teleost fish species (Stevens et al. 2005). At present, fishing pressure on chondrichthyan species is increasing because of the high, and in some cases rising, value of their meat, fins, livers, and/or gill rakers (Fowler et al. 2002, Clarke et al. 2006, Lack and Sant 2009). The landings of sharks and rays reported to the Food and Agriculture Organization of the United Nations increased to a peak in 2003 and declined by 20% afterwards (FAO 2010, Frodella et al. 2016). However, recent global data show an increase of 42% of the trade in shark meat (Dent and Clarke 2015) and the true total catch is considered to be about 3/4 times greater than reported (Clarke et al. 2006, Worm et al. 2013). Most chondrichthyan catches are unregulated and often misidentified, unrecorded, aggregated or discarded at sea, resulting in a lack of species-specific landings information (Iglésias et al. 2010, Bornatowski et al. 2013, Cariani et al. 2017).

Humans have been exploiting marine resources, including sharks, along the Mediterranean coasts since ancient times (Farrugio et al. 1993). An analysis of threat levels across all chondrichthyan species has revealed the Mediterranean as a key hotspot of extinction risk (Dulvy et al. 2014). According to the recent assessment of the International Union for the Conservation of the Nature (IUCN), 50% of rays (16 of 32 species) and 54% of sharks (22 of 41 species) in the Mediterranean are facing a high risk of extinction, whereas the only chimaera species (Chimaera monstrosa) is considered as least concern (Dulvy et al. 2016). As highlighted above, the principal driver of decline and local extinction is overfishing. Most species are taken as retained valuable by-catch in the bottom trawl and the small-scale multispecies fisheries. However, by-catch volumes and species composition of chondrichthyans are poorly documented in the Mediterranean and data are rarely incorporated into national and international statistics. Therefore, sharks and rays caught as by-catch can be only delivered as an approximation (Camhi et al. 1998, GFCM 2014). Moreover, despite being banned in 2002, illegal driftnetting is widespread throughout the basin: fleets from Algeria, Italy, Morocco, Turkey, among others, continue to fish illegally with pelagic driftnets (Notarbartolo Di Sciara 2014), and this represents an important and largely hidden source of mortality for pelagic sharks. More recently, a significant decline in chondrichthyan species richness has been reported throughout the Mediterranean, due to increasing threats and local extinctions (Nieto et al. 2015, Davidson et al. 2016, Dulvy et al. 2016). Starting from 2010, the General Fisheries Commission for the Mediterranean (GFCM) has adopted measures to reduce the by-catch of pelagic sharks, has banned finning practices and has also prohibited the capture and sale of the sharks and rays species listed in Annex II of the Protocol of the Barcelona Convention (FAO 2016). However, no effective management measures against the overexploitation of many commercial species have been successfully implemented or enforced (Vasilakopoulos et al. 2014, Osio et al. 2015). In addition, data on the stock status of sharks and rays remain poor or non-existent (Polidoro et al. 2008, Worm et al. 2013, Colloca et al. 2017).

Numerous studies on chondrichthyan species were made during the first decade of the 21^st century in the western (e.g. Carbonell et al. 2003, Moranta et al. 2008a, b), central (e.g. Sion et al. 2004, Abella and Serena 2005, Ferretti et al. 2005) and eastern Mediterranean (e.g. Megalofonou et al. 2005, Tserpes et al. 2006, Peristeraki et al. 2008). However, only a few of them have been updated, mainly those related to Spanish (Guijarro et al. 2012, Barría et al. 2015, Ramírez-Amaro et al. 2015), Italian (Follesa et al. 2011, Ligas et al. 2013, Marongiu et al. 2017) and Greek waters (Maravelias et al. 2012, Tserpes et al. 2013). These studies have been carried out on the northern Mediterranean coast and very few have been carried out on the southern coast off Africa (e.g. Ordines et al. 2011).

The aim of this work is to give updated information on the distribution and abundance of demersal chondrichthyans on a wide Mediterranean scale and to compare findings among different areas in relation to environmental factors. Abundance indices and frequency of occurrence of the species were estimated from data obtained from the International Mediterranean Trawl Survey (MEDITS), which aims to produce information on population distribution and demographic structure of demersal species inhabiting the continental shelf and the upper slope (Bertrand et al. 2002). Thanks to the adoption of a common standardized sampling protocol, this scientific survey provides comparable species-specific data in several parts of the Mediterranean.

The data were collected during the bottom trawl MEDITS surveys, which have been performed annually in spring-summer since 1994. This research programme involves several institutes and teams across different geographical sub-areas (GSAs) established by the GFCM. More specifically, data of chondrichthyans at species level were provided by 16 GSAs from 9 different countries (Spain, France, Italy, Slovenia, Croatia, Albania, Montenegro, Greece and Cyprus), covering wide areas of the western, central and eastern Mediterranean (Table 1; Fig. 1). Considering that not all chondrichthyan species were included in the list of MEDITS target species until 2008, there were no complete data time series available for all study areas. For this reason, only data of four recent years (from 2012 to 2015) were analysed. For the Greek GSAs (20, 22 and 23), data were available only for the year 2014, as the MEDITS survey was performed only in this year.

As described in the MEDITS Handbook (Anonymous 2017), these surveys were carried out from May to July, with the exception of 2015 when some Italian GSAs postponed the survey to early autumn due to bureaucratic problems. Samples were taken during daylight using the experimental bottom trawl GOC-73, with 10 mm cod-end mesh size, which corresponds to a mesh opening of about 20 mm. This gear was specially designed to be highly selective in all sampled areas and depths, allowing a representative sampling of demersal megafauna species. Sampling station selection was based on a stratified sampling scheme including five bathymetric strata: 10-50 m, 51-100 m, 101-200 m, 201-500 m and 501-800 m, except for GSA 17 in which the deepest stratum was not present. The duration of hauls was 30’ at stations at less than 200 m depth and 60’ at deeper stations. The number and weight of all specimens captured were recorded. For each species, the frequency of occurrence within their preferential depth range was calculated by GSAs as percentage of hauls where it was present in respect to the total hauls of the stratum during the four-year study period (Table 2). Abundance indexes (biomass index as B in kg km^–2 and density index as D in ind. km^–2) were calculated for each species by GSAs for the following depth ranges: overall range (10-800 m), continental shelf (10-200 m) and slope (201-800 m).

Following the last taxonomic structure suggested by Eschmeyer et al. (2018), 41 species of chondrichthyans were detected from all the GSAs. They included 18 sharks (5 orders and 11 families), 22 batoids (3 orders and 4 families) and 1 chimaera (Table 2).

This work provides standardized and updated information regarding the distribution, occurrence and abundance of chondrichthyans along the whole northern Mediterranean, from west to east. Historically, their diversity has been considered greatest in the western basin, particularly in the coastal waters of Morocco, Algeria, and Tunisia (between 57 and 69 species), and slightly lower in the northwestern countries, Spain (including the Balearic Islands), France and Italy, and in Greece. Intermediate levels of diversity have been reported for the central Mediterranean along the coasts of Libya, Malta and Sicily and in the coastal waters of the countries bordering the Adriatic and Aegean seas. This longitudinal gradient in the species richness of chondrichthyans has been suggested by several authors (Dulvy et al. 2016, Melendez et al. 2017, Bradai et al. 2018). However, the results of the present work seemed not to confirm this pattern (the Strait of Sicily and the Aegean showed the largest number of species), probably because North Africa is not involved in the MEDITS programme, so diversity in this area could not be considered.

Comparing our results with the literature, an abrupt decrease in the number of chondrichthyan species has been detected in the northern Adriatic (GSA 17): from 33 demersal, meso-predatory elasmobranchs in 1948 (Jukić-Peladić et al. 2001, Coll et al. 2009, Ferretti et al. 2013) to 20 species found in this study during recent years (2012-2015). This confirms the known structural depletion of the demersal community in terms of diversity during the last few decades.

A previous study conducted by Bertrand et al. (2000) involving almost the same GSAs sa those analysed in this work pointed out that the most widespread species in the basin were Scyliorhinus canicula, Galeus melastomus and Etmopterus spinax among sharks, Raja clavata, Raja miraletus, Raja asterias and Torpedo marmorata among batoids, and Chimaera monstrosa. From our results, the most frequent and abundant species all around the basin are S. canicula, G. melastomus and E. spinax among sharks and R. clavata and R. miraletus among batoids, partly confirming the findings of Bertrand et al. (2000) and other previous data present in literature for the western (Barría et al. 2015, Ramírez-Amaro et al. 2015, Marongiu et al. 2017), central (e.g. Jukić-Peladić et al. 2001, Bottari et al. 2014) and eastern (Tserpes et al. 2013) Mediterranean. The overlapping of observations should probably be attributed to the higher fishing pressure resilience of these species (Cavanagh and Gibson 2007). Usually, in spite of this resilience, fishing pressure should be considered as one of the main variables affecting the abundance of chondrichthyans, but not the only one (Frodella et al. 2016).

To parameterize fishing pressure is a big issue, particularly in terms of spatial distribution and at a large scale (Russo et al. 2017). In fact, although a spatial index of fishing pressure was recently developed at a regional scale in the Mediterranean (Kavadas et al. 2015), the only maps of the geographical distribution of trawling fishing pressure along the whole basin were drawn up by Sbrana et al. (2013). The comparison between these maps and our results shows that the areas under higher fishing pressure, such as the Adriatic and the Spanish coast (with the exception of the Balearic Islands) show low abundance of chondrichthyans, but other areas with a high level of fishing pressure, such as southwestern Sicily, show a high abundance of these species. This finding suggests that other environmental drivers work together with fishing pressure to shape their distribution.

Some differences must be noted between our results and those of Bertrand et al. (2000), especially regarding the decline of C. montrosa. This species seemed not so widespread as in the past years. A possible explanation for this decline, besides its rarity, could be found in the fact that this bathyal species has a lower degree of survival after discarding than other chondrichthyans (Ferretti et al. 2005). Among the rare or non-common species found in our analysis, the order Squatiniformes deserves particular attention. In the last century, angelsharks (Squatina spp.) were common in Mediterranean waters (Ferretti et al. 2005). Fortibuoni et al. (2016) reported that Squatina squatina was commonly sold in the main fish markets until the collapse of this species during the 1960s led to its disappearance in the northern Adriatic. Furthermore, Damalas and Vassilopoulou (2011) underlined the absence of the two other species of angelsharks, Squatina aculeata and Squatina oculata in the Aegean Sea, reporting that they had not been present in the catches for a couple of years. Our study shows that the only species of angelsharks caught during recent years is S. aculeata, whose presence has been detected in the same area. This fact could be a clear sign of decline or indicate a possible risk of extinction of this and the other two congeners in the Mediterranean, although, fortunately, interviews with fishermen reveal that this species seems not to have been extirpated yet, as was also observed in the Tyrrhenian (Ligas et al. 2013).

Some authors have reported the spiny dogfish (Squalus acanthias) to be one of the most frequent shark species captured in the Mediterranean (Baino et al. 2001, Serena et al. 2009, Damalas and Vassilopoulou 2011). Our results show its congener S. blainville to be the most abundant Squaliformes species throughout the basin, particularly in its central-western part (eastern Corsica and southern Sicily) and in the eastern Ionian and Aegean seas. Tsagarakis et al. (2013) also reported S. blainville to be one of the most important species of the demersal assemblages in the eastern Ionian Sea. The reason for this replacement between these two sharks could be related to taxonomic problems afflicting the Squalus genus and currently solved in different areas of the Mediterranean (Bonello et al. 2016, Bellodi et al. 2018). Indeed, recent studies (e.g. Anastasopoulou et al. 2018) highlight that S. acanthias showed a limited geographic distribution in the past, suggesting an inaccurate classification of these two species.

Focusing on macro-areas, previous findings regarding the distribution and abundance of chondrichthyan species conducted in Spanish and French seas (Massutí and Moranta 2003, Gouraguine et al. 2011, Ramírez-Amaro et al. 2015) are, in general, confirmed in the present work, with the highest number of species and abundance around the Balearic Islands. Regarding Italian seas, our data, as in a study conducted by Relini et al. (2000), confirmed the highest diversity of demersal chondrichthyans in the Strait of Sicily. Furthermore, G. melastomus and S. canicula were shown to be the most frequent and abundant shark species in Spanish and Italian waters, and R. clavata and R. miraletus were the most frequent and abundant batoids. In particular, R. clavata, with a large depth range, seemed to be more abundant off the Balearic Islands (GSA 5) and western Italy (GSAs 9, 10 and 11), where the slope is wide. R. miraletus, which is mostly widespread on sandy bottoms of the shelf, is also abundant around the Balearic Islands and off the central part of Italy (GSAs 16, 17, 18, 19), particularly in the northern Adriatic (GSA 17), where it is the most abundant skate. The low depth of this area, where the slope represents only 7% of its surface, explains the scant presence or the absence of the deep species E. spinax and G. melastomus, which are common in other GSAs, as also reported by Ferretti et al. (2013). The presence of the two Hexanchiformes Heptranchias perlo and Hexanchus griseus in Greek waters is probably due to the fact that they inhabit restricted and deep areas in favourable biological environments that are not exposed to a high fishing pressure, such as the Greek seas (Sion et al. 2004, El Kamel-Moutalibi et al. 2014, Rodríguez-Cabello 2018).

The modelling approach explored in this work has allowed us to produce useful maps to describe the distribution of demersal chondrichthyans in the Mediterranean following environmental drivers and mostly in terms of depth and temperature ranges. Within the current context of global warming, these environmental factors can play a pivotal role in the management and conservation of these species. However, taken into account their vulnerability to fishing exploitation, other factors such as fishing pressure should be incorporated in the future to obtain an exhaustive picture of the distribution, occurrence and abundance of chondrichthyan species inhabiting the Mediterranean. Considering their great number of species and abundance along northern African coasts, and the lack of time series of data available from these waters (Dulvy et al. 2016), the research surveys and investigations should be extended to the southern Mediterranean.