Bio-Ecological Features Update on Eleven Rare Cartilaginous Fish in the Central-Western Mediterranean Sea as a Contribution for Their Conservation

Cartilaginous fish are commonly recognized as key species in marine ecosystems for their fundamental ecological role as top predators. Nevertheless, effective management plans for cartilaginous fish are still missing, due to the lack of knowledge on their abundance, distribution or even life-history. In this regard, this paper aims at providing new information on the life-history traits, such as age, maturity, reproductive period, in addition to diet characteristics of eleven rare cartilaginous fish inhabiting the Central-Western Mediterranean Sea belonging to the orders Chimaeriformes (Chimaera monstrosa), Hexanchiformes (Heptranchias perlo and Hexanchus griseus), Myliobatiformes (Aetomylaeus bovinus and Myliobatis aquila), Rajiformes (Dipturus nidarosiensis and Leucoraja circularis), Squaliformes (Centrophorus uyato, Dalatias licha and Oxynotus centrina) and Torpediniformes (Tetronarce nobiliana), useful for their assessment and for future management actions. Particularly, the present paper provides for the first time the age estimation of D. nidarosienis and L. circularis which were both found capable of becoming older than ten years. In addition, the present study updates the sizes of first maturity of C. uyato and D. licha, which appeared to be capable of reproducing earlier than what was previously hypothesized, representing very valuable information for a better understanding of these rare species populations status and, eventually, their conservation. On the basis of the stomach content analysis, it was possible to identify five different predator groups.


Introduction
Marine ecosystems represent Earth's largest biodiversity reservoirs [1], providing a wide range of contributions to human well-being, which collectively go under the name of Marine Ecosystem Services [2,3]. Due to this high resource availability, the human pressure on marine ecosystems is constantly rising, thus negatively impacting their biodiversity [4][5][6]. In this context, the Mediterranean Sea is considered, at the same time, a biodiversity hotspot and one of the most threatened marine areas in the world [7][8][9][10]. Even though the Mediterranean basin represents only 1% of the global oceans, it is estimated that it hosts up to the 18% of the world's macroscopic marine species, about 30% of which are endemic [7]. Among these, cartilaginous fishes such as sharks, batoids (skates and rays) and chimaeras are certainly some of the most iconic species for the majority of public opinion, mainly due to their fundamental ecological role as top predators [11,12]. At the same time, due to their k-selected life-history traits, cartilaginous fish are particularly vulnerable to overfishing, habitat deterioration and to the modifications in the biological communities' All specimens were weighed (Total Mass, TM, in g) and the main biometrics were recorded as follows: Total Length (TL, in cm) was used for the species belonging to the Torpediniformes, Rajiformes, Hexanchiformes and Squaliformes orders, while Disc Width (DW, in cm) and Pre-Ventral Length (SVL, in cm), were used, respectively, for All specimens were weighed (Total Mass, TM, in g) and the main biometrics were recorded as follows: Total Length (TL, in cm) was used for the species belonging to the Torpediniformes, Rajiformes, Hexanchiformes and Squaliformes orders, while Disc Width (DW, in cm) and Pre-Ventral Length (SVL, in cm), were used, respectively, for Myliobatiformes and Chimaera monstrosa due to the fragile nature of the filamentous tail that characterize these species. Length-Mass relationships were calculated for the combined sexes according to the power curve function M = aL b , where M represents the Total Mass (TM) in g and L the main length used in cm (TL, DW and SVL depending on the species), a represents the intercept of the regression and b is the regression coefficient. The eleven studied species are shown in Figure 2. Myliobatiformes and Chimaera monstrosa due to the fragile nature of the filamento that characterize these species. Length-Mass relationships were calculated for the bined sexes according to the power curve function M = aL b , where M represents the Mass (TM) in g and L the main length used in cm (TL, DW and SVL depending species), a represents the intercept of the regression and b is the regression coefficien eleven studied species are shown in Figure 2.
For the most abundant species, ovarian fecundity, defined as the total number of yolked eggs, was estimated by counting the eggs in both ovaries of mature females (stages 3a-4a). Uterine fecundity, defined as the number of pups, was also recorded in females at mid-and late-term pregnancy (stages 3c-3d).

Age Estimation
In order to estimate specimens' age, a portion of the vertebral column, including 8-10 centra, was extracted from the thoracic region of all species, with the exception of A. bovinus, M. aquila and C. monstrosa. To remove the exceeding soft tissues or muscles, after the removal of neural and haemal arches using a scalpel, each centra was soaked in a 5% hypochlorite solution for 5-10 min [29], depending on the centra dimension. With the aim of identifying the method that could allow the best band visibility, from a subsample of at least 30 individuals (15 females and 15 males of different size classes), 3 different centra per individual were selected and treated with two of the most common staining methods known in literature, Alizarine Red [30] and Silver Nitrate [31,32], and tested against the third centra kept unstained. The ageing process was carried out on vertebral centra sagittal sections following [33]. The section final thickness of 0.5 mm (±0.1 mm) allowed a better growth band visibility [34]. Vertebral sections (thickness 0.5 mm ± 0.1 mm) were obtained by grinding the whole centra embedded in a bi-component epoxy resin (Bio Optica, Technovit EPOX), grinding them using a polisher (ATM Saphir 320) equipped with progressively thinner abrasive discs [34,35]. Finally, age was obtained following the ageing methodology proposed for Elasmobranchs by [33].

Stomach Content Analysis
Stomachs were dissected and stored in a 5% formaldehyde solution. Then, contents were analyzed and classified to the lowest taxonomic level possible. The percent Frequency of Occurrence (%FO) of a specific prey item in all samples was calculated [12].
Data on D. nidarosiensis, collected in the period 2005-2011, and on T. nobiliana, collected between 2008 and 2019, have been presented in [36,37] and [38], respectively, while this study aims at updating the available information up to 2021.

Leucoraja circularis
In the period between 2008 and 2020, 77 sandy ray specimens were found in sea bottoms from 128 to 620 m (mean ± s.d. = 382 ± 140.8 m) (Figure 1e). Specimens' size ranged between 17.1 and 90.3 cm in TL (mean ± s.d. = 46.1 ± 16.1 cm) ( Table 1) with weights of 17 to 3917 g (TM) (mean ± s.d. = 711 ± 922 g). Length-Mass relationship was M = 0.001 × L 3.360 (r 2 = 0.99) ( Table 1). Conversely to what was observed in some of the other analyzed species, females and males presented no major difference in size (K-S, p > 0.05). Females (N 42) ranged between 17.1 and 90.3 cm TL (mean ± s.d. = 46.9 ± 17.6 cm), while males' (N = 35) TL varied between 19.8 and 73.4 cm (mean ± s.d. = 45.0 ± 14.4 cm) ( Figure 3d). However, females were found able to reach much higher weights than males, at 17-3917 g (TM) (mean ± s.d. = 631 ± 751 g) and 25-1872 g (TM) (mean ± s.d. = 541 ± 494 g), respectively. The analysis of vertebral centra thin sections (Figure 4b) revealed the species as characterized by a potentially medium/long life span. The oldest male and female individuals (TL = 73.4 and 76.6 cm, respectively) were both estimated to be 11 years old. Unfortunately, it was not possible to estimate the age of the biggest individual caught (90.3 cm TL). Only two of the caught specimens, both collected in June, were found mature, precisely corresponding to the biggest female (90.3 cm TL) classified as capable of reproducing, and to the biggest male (73.4 cm TL) which was in regressing stage. The prey items found in the 19 full stomachs belonged to four main groups: Crustaceans, which represented the most important (%FO = 63), followed by Actinopterygians, other Elasmobranchs and Cephalopods (%FO = 53, 21 and 21, respectively).

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Myliobatis aquila
Similar to the bull ray, the common eagle ray also showed the narrowest bathymetric distribution among the analyzed species, being exclusively found in shallow coastal waters in the depth interval between 29 and 43 m (mean ± s.d. = 34 ± 3.0 m) (Figure 1c). The size of the 91 collected individuals ranged between 25.2 cm and 58.7 cm in DW (mean ± s.d. = 36.0 ± 73.6 cm) (Table 1), with TM between 220 and 5870 g (mean ± s.d. = 878 ± 920 g). Length-Mass relationship was M = 0.011 × L 3.076 (r 2 = 0.93) ( Table 1)

Torpediniformes Tetronarce nobiliana
During the period between 2009 and 2021, 26 great torpedo ray individuals (13 females and 12 males) were captured. The specimens were caught at depths between 200 and 600 m (mean ± s.d. = 489 ± 127.8 m) (Figure 1f). While the minimum observed size recorded was similar for females (TL = 24.2 cm; TM = 222 g) ( Table 1) and males (TL = 24.0 cm; TM = 211 g) (Table 1), the biggest female (TL = 101.5 cm) was nearly twice the size (TL = 66.6 cm) and weighed almost three times more (female TM = 15130 g versus male's TM = 48 g) than the biggest male. The mean sizes (±s.d.) recorded were 44.9 ± 25.7 cm (TL) and 2830 ± 4923 g (TM) for females and 44.6 ± 14.5 cm (TL) and 1842 ± 1550 g (TM) for males (K-S, p > 0.05; Figure 3e). Length-Mass relationship, calculated for the mixed sexes, was M = 0.014 × L 3.026 (r 2 = 0.99) ( Table 1). The direct age estimation, carried out on sectioned vertebral centra (Figure 4c), revealed that females were capable of growing older than males, attaining 17-18 years and 9 years, respectively. Only two mature females were found, one was recognized as capable of reproducing (TL = 98.0 cm, caught in January), the other as regressing (TL = 101.5 cm, caught in April). Concerning males, three mature specimens were collected, two of them were capable of reproducing (TL = 66.6 cm caught in July and TL = 59.3 cm caught in April) while the other was active (TL = 54.0 cm, caught in February). Actinopterygii (%FO = 100) represented the only prey item found in T. nobiliana; the presence of remains of bony fishes were recorded in every analyzed stomach.

Discussion
The present study aimed to enrich the available literature on rare and poorly-known cartilaginous species, also taking advantage of the relatively high abundance of these species in Sardinian waters [17,24], providing information on some of their life-history traits, such as age, maturity, reproductive period, in addition to diet features. This kind of information has also been recognized as extremely useful in consideration of the global IUCN Red List reassessment, which recently started from the Northeast Atlantic and Mediterranean Sea cartilaginous fish species [18]. Indeed, taking into account the extreme data deficiency and general lack of information characterizing these species, this new assessment was highly focused on trait-based predictions using biological and ecological trait data [18]. For example, the main traits associated with a higher extinction risk were large body size, small bathymetric and geographic range, and ecological specializations (such as a narrow trophic niche) [18]. With respect to the existing data on the species analyzed, this work represents a useful update toward a more comprehensive assessment and, consequently, an improvement for their conservation status.

Bathymetric and Geographic Distribution
Some bathymetric ranges have been updated with respect to what is reported in literature. The species belonging to the Myliobatiformes order, for example, exhibited a much narrower bathymetric distribution in Sardinian waters, being always found in the first 50 m of the water column, while in literature their range is extended up to 150 m for A. bovinus [39,40] and to 573 m for M. aquila [41]. This fact could indicate, for these species, Life 2021, 11, 871 12 of 16 a higher degree of vulnerability than previously estimated, particularly at the regional level. The shallower waters are, indeed, the most exploited by artisanal fishery and in general represent the marine areas most impacted by a large amount of anthropic activities. In this regard, presenting such a narrow bathymetric range limited to this particular area could represent an additional level of threat for these species' conservation status. Despite the relative rarity of C. monstrosa, 47 of the total 192 individuals were caught together in a single commercial fishing haul that occurred on 24 July 2020 (38 • 50 N; 9 • 38 E). Most of these specimens were immature and quite small (mean SVL = 13.1 cm). This peculiar aggregation might indicate the presence of an essential habitat for the species, such as a juvenile feeding ground or even a nursery area. Unfortunately, the data collected during a single haul do not consent to verify properly the latter assumptions. However, considering that the designation of essential fish habitats is one of the main goals that environmental managers are facing [42,43], and given the obvious implications for the species conservation, further studies must be endorsed to investigate if the area may represent a sort of sanctuary or a nursery ground for the species sensu [44].

Size Range
According to [45,46] the maximum reported size for H. griseus was 482 cm (TL) even though both studies hypothesized that the species can probably reach Total Length of about 550 cm. In this regard, the observed size of the biggest specimens caught in Sardinian Seas (518 cm TL) seems to endorse the hypothesized maximum size proposed by these authors. In consideration also of the records of two large specimens described by [47], with TL 523 cm and 600 cm, caught in the South-East Aegean Sea and the Sea of Marmara, respectively, the largest Sardinian specimen appears to be the third largest bluntnose sixgill shark ever recorded and consequently the largest in Western Mediterranean Sea.
A strong size-related sexual dimorphism (SSD), with females attaining bigger sizes than males, is quite often reported in sharks (e.g., [37,[48][49][50][51]) and batoids (e.g., [35,52,53]). Most of the species analyzed in the present study seemed to follow this trend, except for both eagle ray species M. aquila and A. bovinus, which exhibited no signs of SSD, representing an apparent exception in Cartilaginous fish biology. Certainly, the low sample number, especially regarding A. bovinus, may have affected this outcome. However, considering that similar findings have been reported for other Myliobatiformes (e.g., the white-spotted eagle ray Aetobatus narinari, [54]), the hypothesis of no SSD in these Mediterranean species seems to be plausible.

Age and Growth
Information on a certain species age and growth are generally considered essential to a proper assessment [53]. In this regard, the present paper provides the very first age data, although preliminary, on the D. nidarosiensis and on L. circularis since, accordingly to [55] for the latter species, "Age at maturity, longevity, size at birth, reproductive age, gestation time, reproductive periodicity and fecundity are unknown". Growth bands on both species vertebral centra thin sections appeared relatively easy to be recognized, although in D. nidarosiensis, similarly to its congeneric D. oxyrinchus [35], the presence of growth bands split in doublets or triplets have been detected. For this reason, the age estimation on this species could be tricky and must be carefully evaluated. Both skates' species seemed capable of reaching relatively old ages, and the maximum observed age in L. circularis appeared highly consistent with the longevity reported for its congeneric L. naevus [56]. The relatively high longevity showed by these skates contributes to indicating them as species particularly prone to conservation risks, sensu [18], thus highlighting the need to further investigate this aspect, considering also the high reliability of the age estimation protocol employed in this study [33,57].
Indeed, the age estimation analysis showed vertebral centra thin sections as a highly consistent and easily interpretable structure for ageing purposes on the analyzed species belonging to the Rajiformes and Torpediniformes orders. Conversely, the ageing protocol was not able to allow any attempt of age determination of the deep-sea sharks belonging to the Hexanchiformes and Squaliformes orders, since none of the applied staining methods permitted to enhance the contrast between growth bands. This issue appears to be related to the extremely low calcification level of this species skeleton, already documented in species such as H. griseus [58]. Considering the important role that age and growth data cover in species management and conservation and the species-related effectiveness of staining methods [29,58], future studies testing new methods on poorly-known species should be endorsed. Information regarding species age and growth are generally recognized as difficult to achieve, especially for cartilaginous species, which often present additional issues with respect to bony fish [38]. One of these problems is the lack of age validation studies, such as mark-recapture campaigns that are extremely subject to a variety of issues, e.g., tagging-related mortality, tag loss and tag detection and reporting rates. These problems are further worsened by the low recapture probability, linked to the rarity of the species [57]. For this reason, most of the studies did not validate the band deposition periodicity in vertebral centra and even fewer the absolute age [57]. The majority of the studies, indeed, are only based on the annual patterns of growth areas in the form of a succession of opaque and transparent bands that are assumed to be annuli, as in the present work. In this regard, given the data paucity characterizing these species, every kind of information results are crucial.

Reproductive Features
Concerning reproductive features, the present paper updated the available information on some of the analyzed species, such as the reproductive period and the size of the smallest mature individuals. Regarding the Hexanchiformes order, an active H. perlo male smaller than the reported first maturity size [21] has been found in the present paper. Additionally, mature males below the reported first maturity size have been observed also in the Squaliformes C. uyato, whereas this situation in D. licha emerged in both sexes [21]. Apparently, the present study updated the sizes of first maturity of these species. All of them appeared to be capable of reproducing earlier than what was previously hypothesized. Although, due the rareness of some of the analyzed species a rather scattered observation pattern was returned, generally, a continuous breeding cycle throughout the year can be hypothesized. The last statement seems to fit particularly well to the analyzed species that inhabits deeper water, confirming the reproductive pattern of deep-water cartilaginous fishes commonly reported, e.g., [34,51] and references therein. These results may represent very valuable information for a better understanding of these rare species' populations status and, eventually, their conservation.

Diet
The stomach content analysis showed that the studied species could be gathered in five different groups on the basis on their main prey items. The first group, constituted by Myliobatiformes, was defined by the high abundance of gastropods and hermit crabs in the diet. It appeared interesting to note how, despite the peculiar mouth structure of eagle rays, particularly adapted in crushing shells, our results showed A. bovinus to feed mostly on soft-bodied gastropods, such as Opisthobranchs. This result could represent for the species a strategy of avoiding competition with M. aquila that share the same habitats. The second group included the Rajiformes and C. monstrosa, which seemed to share a preference for Crustaceans, showing sometimes a high preponderance for prey, such as the large deep-water crab G. longipes in D. nidarosiensis. Our data represent the very first information on the L. circularis diet and a huge update on the Norwegian skate in the Mediterranean Sea. T. nobiliana constituted another separated group; its sit and wait predatory habit, in addition to its capacity to stun even large preys, led to a high importance of bottom-related bony fishes in the diet [38]. The fourth group comprised deep-water sharks belonging to Hexanchiformes, both H. perlo and H. griseus, and the Squaliformes D. licha and C. uyato, which fed on relatively large and fast-moving preys such as bony fishes, squid-like Cephalopods and, in D. licha and H. perlo, other Elasmobranchs. Finally, O. centrina's feeding habits seem focused on other Elasmobranchs' egg capsules, due to their stomachs only having the presence of yolk-like liquid and other Elasmobranchs' embryos. These findings appeared in disagreement with the worm-eating description of this species proposed by [59], but in line with [60], who identified O. centrina as an egg-case predator. In this regard, new studies based on molecular analysis are needed to clarify this peculiar aspect. In addition to the large body size and bathymetric distribution [18], particular feeding habits could also play an important role in defining species vulnerability.
In conclusion, the present paper provided new, and in some cases the first, information regarding reproduction, growth and feeding behavior of eleven rare and poorlyknown Elasmobranch species. This kind of information represents a solid step toward a better comprehension of the life-history of these species and, given the unavoidability of these data for any effective management plan, an extremely valuable baseline for future conservation action on these key species for ecosystem equilibrium.
Author Contributions: A.M. and A.B. equally contributing authors, conceptualization, methodology, investigation, formal analysis, writing-original draft preparation, investigation; investigation, review and editing, A.C.; methodology, formal analysis, investigation, writing-review and editing, P.C., M.F.M. and C.P.; data curation, investigation, review and editing, P.P.; supervision, project administration, review and editing, M.C.F. All authors have read and agreed to the published version of the manuscript.

Funding:
The work has been funded by the European Union and the Italian Ministry of Agricultural, Food and Forestry Policies, in the framework of the Italian Work Plan for data collection in the fisheries and aquaculture sectors.