Using the last complete annotated checklist from the region (Santos et al., 1997) as our starting point, we reviewed the list of Chondrichthyes (Elasmobranchii and Holocephali) found in the Azores EEZ. Additions and deletions were made strictly using primary scientific data, where the species were identified directly by its authors. This included results from experimental fisheries surveys, underwater visual surveys and other published literature describing the biodiversity within the Azores EEZ. The results were compared with the FAO Fisheries Catalogue for the region (Ebert and Stehmann, 2013) as a measure of global knowledge about the region. Clarifications regarding taxonomy and species occurrence were sought from local experts where necessary.

The primary sources used were:

Scientific fishing surveys using bottom longlines carried out by the Department of Oceanography and Fisheries, University of Azores (DOP/UAz), Horta (Menezes et al., 2006, 2012; Menezes and Giacomello, 2013; Menezes, 2014) around the nine islands and important seamounts in the Azores EEZ;

Species were classified into four classes to designate their presumed frequency of occurrence. The different classes were differentiated by color codes as follows:

Commonly encountered species were designated green—these species have multiple records and no known taxonomic conflicts;

The updated species list was then used to analyze the biogeographical relationship of the Azorean elasmobranch assemblage within the North Atlantic. In order to maintain analytical coherence through a broad comparable environmental envelope, we focused our analysis on the section of the Atlantic bounded by the 42°N (northern extent of Azores EEZ) and 26°N (south of the Canary Islands) latitudes, thus including the Macaronesian archipelagos of Madeira and the Canaries, and the continental margins of the Atlantic to the east (Iberian Peninsula) and west (United States east coast). This section of the North Atlantic was assumed to represent the highest potential biogeographic affinity of species present in the Azores. Tropical regions such as, the archipelago of Cape Verde and Caribbean islands were excluded being separate biogeographic regions (Floeter et al., 2008) and owing to the few highly-migratory species shared between these sections of the North Atlantic.

Species lists of the other regions were adapted from regional annotated checklists (Brito et al., 2002; Wirtz et al., 2008; Wirtz, 2011) and recently published literature [Freitas and Biscoito, 2007; Ferreira et al., 2008; Freitas et al., 2011; Kyne et al., 2012 (only species from NWA); Ebert and Stehmann, 2013; Delgado et al., 2017], augmented by local expert knowledge where possible (P. Pascual IEO/Canaries, pers. comm. 2017; M. Freitas OOM/Madeira, pers. comm. 2017). Updated, comprehensive, region-specific annotated checklists were often not available, leading to a certain probability of erroneous species records. However, these were not considered to significantly affect the overall biogeographic pattern.

First, we focused on the elasmobranch species occurring in the Azores and how they are shared with the study areas across the North Atlantic. We added current knowledge on migratory habits of these species (Fowler, 2014) to overlay the observed biogeographic patterns. Next, we built a presence-absence table of all species by region, and analyzed biogeographic similarities using cluster analysis. The dissimilarity matrix was constructed with binary distances, and unweighted agglomerative method (UPGMA) used to build the dendrogram. All calculations were executed in RStudio® software environment using the “base” package. The structure of elasmobranch assemblages divided by family was used to understand which species groups influence similarities and differences between regions.

To determine threats to elasmobranchs at the regional level, we focused on the landing records of local fisheries; since there is no clear evidence of habitat destruction, pollution, or other factors threatening the local elasmobranch assemblage. The goal was to identify gear types more likely to land elasmobranchs and the species landed by these gears. Analysis of the landing records also provides the opportunity to recognize potential inaccuracies in the registration process itself. Local fisheries began recording elasmobranch landings in greater taxonomic resolution in official statistics since the mid-1990s. Thus, we chose a 19-year period from 1996 until 2014 to analyze how the local fishery interacts with elasmobranchs, using temporally aggregated landing information.

Landing records are collected from the auction houses (lotas) or point-of-first sale from all nine islands of the archipelago. It is maintained by Lotaçor SA in collaboration with the Department of Oceanography and Fisheries of the University of the Azores. The dataset used in this study was extracted from the Lotaçor database, structured by year, month, gear type (“métier”), and weight landed (kilograms) of each landing within the study period. A subset of this data included the weight of elasmobranch species per landing. Elasmobranchs appeared in the database under their local names with corresponding documentation on scientific names (Lotaçor, 2015).

Having no direct information on catch or effort, we scaled the impact of different métiers using two indicators. The first indicator was the weight of elasmobranchs landed proportional to other species by each métier. The weight of non-elasmobranch species was used as proxy for effort since they are often the target or more lucrative species. Thus, métiers with higher proportional weight landed of elasmobranchs, i.e., weight per unit effort (WPUE), could be preferentially landing these species. The second indicator was the frequency of elasmobranch landings per métier, measured as a percentage of total landings. Using the total number of landings per métier as a proxy for effort, métiers with higher percentage of landings with elasmobranchs, or landings per unit effort (LPUE), would be those more partial to these species.

WPUE was calculated by aggregating the total weight and elasmobranch weight landed by each métier for every month in the study period, separately. The weight of other species was obtained by subtracting weight of elasmobranchs from the total weight, and the ratio obtained for landed weight of elasmobranchs to other species. The arithmetic mean of the ratios was used to reduce the influence of disproportionately large values, and standard deviation used as an indication of variability. Similarly for LPUE values, the total number of landings per métier and the number of landings with elasmobranchs were aggregated to obtain the frequency of elasmobranch landings. All calculations and data analysis was executed in RStudio®.

The métiers were then separated into two categories to facilitate meaningful comparisons. One category featured métiers with higher elasmobranch LPUE (both proportional weight and landing frequency), thus greater tendency to catch and land these species, and a second category for métiers that land less elasmobranchs. The species composition of the two groups was analyzed separately. We used average weight of species per landing to identify species that are landed infrequently but could require greater management attention.

The species landed by the local fisheries were then related to their regional IUCN Red List (Nieto et al., 2015) status as a proxy for potential conservation concern. This analysis was then complemented with an evaluation of information collected opportunistically and mostly available through gray literature or unpublished data on known and putative essential fish habitats (EFH), to preliminarily identify potential areas of greater overlap with fisheries.

Our literature search confirmed 61 chondrichthyans from the Azores EEZ (39 sharks, 17 batoids, and 5 chimaeras; Table 1), of which four species still require taxonomic clarification. The previous annotated checklist (Santos et al., 1997) included 44 species identified from the Azores. Thus, our study adds 19 species (40%) to that list and removes two species following taxonomic corrections. Six species presumed to commonly occur in the Azores did not feature or were considered uncertain in the earlier checklist. We compiled an annotated checklist of the species that do not appear in the main checklist of Santos et al. (1997), which is presented here. The species for which a declared first record could not be found were marked with an asterisk (^*).

Since the distribution of C. opalescens, B. cf. centroura and D. cf. intermedius are still uncertain, the remaining 58 species were used to compare biogeographic affinities of the Azores with the rest of the North Atlantic. About one-third of elasmobranch species of the United States east coast occur in the Azores, while the archipelago shares more than half of the species of continental Europe (Figure 1). Elasmobranch species composition in the Azores was dominated by non-migratory species, majority of which were same as eastern North Atlantic. Migratory species that occur in the Azores were evenly present across the North Atlantic, with a slightly lower number of these species occurring in the Canaries.

The elasmobranch assemblage was most similar between the Azores archipelago and Madeira (Figure 2), with Madeira sharing almost three-quarters of the species with the Azores (Figure 1). The diversity of elasmobranch families of these two regions was comparable as well (Figure 3). Deepwater Squaliformes dominated in species diversity in both regions. However, Madeira had a greater number of carcharhinids while Azores had a higher count of lamniforms.

Not all archipelagos of the Macaronesian biogeographic unit clustered together; the Canaries and Portugal continental shelf formed a separate group (Figure 2). The species diversity (Figure 1) and composition of elasmobranch families (Figure 3) of the Canaries was more similar to mainland Portugal than to Madeira and Azores, even if family diversity was lower than the continental margin. Other evident points of difference were the greater diversity of families in the order Carcharhinidae and dominance of rajids on the continental margin as compared to the Canaries.

There was a clear separation between the elasmobranch species composition of the western and eastern halves of the North Atlantic (Figure 2). Elasmobranch assemblage structure by family indicated that the two continental margins had high, but contrasting species diversity. The western margin was dominated by requiem sharks (Carcharhinids) and a high diversity of Myliobatiformes (Dasyatidae to Urotrygonidae) (Figure 3). On the contrary, the eastern continental species assemblage was dominated by Rajiformes. Deepwater species also differ visibly between the eastern and western halves of the North Atlantic. In particular, Squaliformes (Centrophoridae to Squalidae) had strikingly low species diversity in the west compared to the central and eastern parts.

The species composition of the chondrichthyan assemblage around the Azores is dominated by species that are either oceanic or deep-sea, reflecting the dominant marine habitat surrounding the archipelago. The fact that they are underwhelmingly known becomes evident in the lack or misrepresentation of species occurrence records in regional and global species catalogs. However, the Azorean elasmobranch assemblage emerges as more diverse than expected. Most of the new species additions are a result of nearly two decades of additional sampling effort and the use of different sampling gears. Species like the frilled shark C. anguineus and broadnose chimera Rhinochimaera atlantica, for example, were only obtained in experimental deep-water trawl nets (Melo and Menezes, 2002). More importantly, technological advancements over the last two decades and an increasing interest in the deep ocean (Costello et al., 2010) have allowed sampling at greater depths in more remote locations. Nevertheless, better known and accessible areas like coasts and nearby seamounts also revealed new additions. Evidence shows that morphologically similar species have been misidentified due to prevalent assumptions. For example, coastal stingrays are often identified as the common stingray D. pastinaca. A similar looking species, until now considered to be roughtail stingray D. centroura (Afonso et al., 2013), also occurs in the region. Recently, this species has undergone a reclassification to two different species (Last et al., 2016b), prompting the need to re-identify and reconsider the relative abundance of the two morphologically similar species present in the Azores EEZ. A similar story occurred for the devil rays M. tarapacana and M. mobular, and was resolved to yield unexpected results (Sobral and Afonso, 2014).

Continental margins are often richer in biodiversity because of their greater productivity and older geological age. It is thus not surprising that the elasmobranch species on the Azores archipelago are less diverse compared to the continental margins of the Atlantic, a trend common to bony fishes as well (Santos et al., 1995, 1997; e.g., Almada et al., 2013). Despite the low diversity, the Azores elasmobranch species assemblage have a wider provenance, especially when compared with the other Macaronesian archipelagos. This result may be explained by the fact that the Azores region is considered to represent a transition zone (ecotone) between cold and warm temperate waters, where the Gulf Stream provides a warmer input than would be normal at this latitude. This transitional signature explains why half the demersal fish (sensu lato, including elasmobranchs) assemblage is composed of species of subtropical origin (Menezes et al., 2006), and which also account for nearly 90% of the more rare fish occurrences (Afonso et al., 2013). Our results provide renewed support to this scenario by showing that the region falls in the boundary between the southern limit of cold-water species like deepwater skates (genus Bathyraja) and Greenland shark Somniosus microcephalus, and the northern limit of tropical and sub-tropical species like whale shark Rhincodon typus, sicklefin devil-ray M. tarapacana or round sting ray Taeniura grabata. The importance of this region as a fringe habitat is bound to increase with possible shifts in global distribution (Afonso et al., 2013), thus highlighting the need for continued monitoring of the elasmobranch species assemblage around the Azores EEZ.

Biogeographic studies showed that the Azores have greater proportion of coastal fish species from outside north-east Atlantic compared to the entire Lusitanian province (Almada et al., 2013). Shallow water fauna of the Azores also constitutes species with origins in the east as well as the west of the Atlantic basin (Santos et al., 1995). Similarly, the elasmobranch diversity in the Azores also claims different regional affiliations in the North Atlantic. Highly migratory sharks like blue and mako sharks travel from the eastern coast of North America to the Mid-Atlantic Ridge, close to the Azores (Vandeperre et al., 2014a, 2016; Queiroz et al., 2016). The Azores and associated Mid-Atlantic Ridge maybe an important migratory corridor and feeding ground for some pelagic elasmobranch species (Gore et al., 2008; Wögerbauer et al., 2015; Doherty et al., 2017). Conversely, the biogeographic pattern of non-migratory deep-sea sharks and rays seems to highlight the potential barrier posed by abyssal plains, as well as the possible role of seamounts as dispersal “stepping stones.” Comprehensive genetic and telemetry studies can shed light on the role of the wider Azores region in connecting both sides of the Atlantic basin for species of different habitat ecology.

Local fisheries in the Azores operate at a much smaller scale compared to mainland Portugal and other European fleets, where elasmobranch landings can be almost 20 times greater by weight (Correia et al., 2016). However, the interaction between elasmobranchs and the fishery is far from simple. A recent study of elasmobranch fisheries in the Azores identified four main species–tope shark, thornback ray, blue shark, and short-fin mako shark–to be important to the local fishery (Torres et al., 2016). This study overlooked other elasmobranch species that do not dominate official landing records by weight. Our results show that the local fishery in fact involves other species, landed sporadically but in considerable volume. These less frequent species demand greater management attention due to their higher extinction risk and, perhaps more importantly, high discard potential.

Based on our results we propose three broad areas of further research that can help to further determine the importance of the wider Azores region for elasmobranchs.

The first data-gap that requires closer inspection are errors and omissions in local fishery data. Robust estimates on the extent of unreported/misreported and discarded catch are required to gauge the actual impact of the local fishery on regional stocks of sharks and rays. Since landings cannot provide a holistic overview of local population status, sentinel or fishery-independent surveys are also imperative to monitor the local elasmobranch populations. Some of these could use non-invasive techniques such as, baited remote cameras for coastal nurseries and deepwater sharks on slopes and seamount summits. Additionally, fisheries studies would benefit from an analysis of socio-economic incentives for landing or discarding of different species. This could provide clues regarding future exploitation trends. The second major gap is the lack of region-specific studies focused on the biology and ecology of local elasmobranchs. Finally, and perhaps most importantly, the essential habitat function of the Azores as nursery, mating, migration, feeding needs to be studied in detail, and across species and habitats. Resolving the importance of the Azores in the wider context of the Atlantic will need many more studies on connectivity of shark and ray populations with, or isolation from, other parts of the species distribution ranges. Population dynamics and tagging studies using electronic, genetic or chemical markers, for example, together with the continued monitoring of species occurrence, will further elucidate the importance of the region as an ontogenetic, permanent, or transitionary habitat, and help pinpoint critical habitats for eventual protection as well as future species range shifts in light of current climatic change scenarios.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.