Species-specific biological data are commonly combined with fishery operational data to model exploited stocks and determine appropriate levels of exploitation. However, this approach to fishery management is predicated on the ability of fishery data collectors to correctly identify exploited species. Sharks and other elasmobranch fishes (Elasmobranchii) can be particularly difficult to identify in the field, due to the close physical similarity to other species and/or a lack of taxonomic resolution in some lineages. This paper provides an overview of the difficulties surrounding the field identification of a number of elasmobranch species found in the Arabian/Persian Gulf, where they are heavily exploited and in urgent need of careful management.
Although the ecosystem-based approach to fisheries management is becoming increasingly popular, particularly in developing nations, single-species management has been the most common approach to ensuring the sustainable exploitation of fishery resources (Worm et al., 2009). Whereas the former takes a holistic approach to the management of the marine environment (Pikitch et al., 2004), the latter requires species-specific biological data that are commonly combined with fishery operational data to model exploited stocks and determine suitable management measures (e.g. Hart and Reynolds, 2002). The fundamental requirement of this form of management is that the collected data are indeed from the target species, and they are not contaminated with data from non-target species. The impact of any such contamination will depend on the extent of the contamination and on the magnitude of the biological differences between the target and non-target species.
Taxonomic confusion in fisheries can arise for a number of reasons, but these can be grouped into two broad categories, namely (1) the existence of sympatric species of similar physical appearance, most commonly congeners, and (2) the existence of cryptic or undescribed species. Dealing with the first issue is, in most cases, relatively straight forward; as long as field staff are adequately trained and provided with species identification guides that allow for the unambiguous differentiation of congeners via morphometric and/or meristic characters, taxonomic issues can be circumvented. Cryptic/undescribed species, on the other hand, are potentially more problematic as, by their very nature, their existence goes unnoticed. Advances in molecular-assisted taxonomy in recent years have greatly assisted researchers in the flagging of cryptic/undescribed species (e.g. Hebert et al., 2004; Ward et al., 2005; Hubert et al., 2008; Coulson et al., 2011), but such techniques necessitate a proactive approach.
The purpose of this paper is to review the elasmobranch fauna (sharks and related species) of the Arabian/Persian Gulf (hereafter referred to as the Gulf) in the context of potential species identification issues for fishery managers. As is the case throughout most of the world, elasmobranchs are heavily fished by Gulf countries, and the current levels of exploitation are unlikely to be sustainable (Jabado and Spaet, 2017). Reducing this pressure on regional stocks will require a concerted effort by the countries involved, but the difficulties involved in collecting reliable field data must first be addressed.
Shark landings in the Gulf are diverse but dominated by small-bodied carcharhinids (Carcharhinidae) such as the Milk Shark R. acutus, Grey Sharpnose Shark R. oligolinx, Sliteye Shark Loxodon macrorhinus, and Hardnose Shark Carcharhinus macloti (Moore et al., 2012a; Moore et al., 2012b; Jabado et al., 2014; Jabado and Spaet, 2017). Although all of these species exhibit useful diagnostic features (Compagno et al., 2005), they are of similar size and general appearance. Furthermore, they are commonly landed together in large numbers, so vigilance is required when species-specific data are being recorded. The issue is further complicated by the fact that these landings also commonly contain juveniles of larger carcharhinid species such as the Spottail Shark Carcharhinus sorrah, Blacktip Shark Carcharhinus limbatus, and Spinner Shark Carcharhinus brevipinna (Figure 1). Two small-bodied carcharhinid species that pose a particular challenge are the Whitecheek Shark Carcharhinus dussumieri and Human’s Whaler Shark Carcharhinus humani; the former being a commonly landed species in the region (Moore, 2012a) and the latter having only been established as a separate species relatively recently (White and Weigmann, 2014). To date, C. humani has yet to be included in any widely-published shark identification guide, and some fishery researchers in the region may be unaware of its existence.
The potential for misidentification also exists among the adults of larger carcharhinids. Carcharhinus limbatus and C. brevipinna are sympatric throughout much of their respective ranges, including the northern Indian Ocean and its associated gulfs (Compagno et al., 2005), where they are among the most commonly landed large sharks (Henderson et al., 2007; Jabado et al., 2014). Differentiation of these species using taxonomic keys has proved problematic in the past (Clark and Von Schmidt, 1965); an issue that was addressed by Branstetter (1982) in a detailed comparison of their morphometric and meristic characteristics. However, only a single morphometric ratio exhibited non-overlapping values between these species (Branstetter, 1982), underlining the extent of their physical similarity. Other large-bodied, black-tipped carcharhinids within the Gulf, such as the Graceful Shark Carcharhinus amblyrhynchoides are also likely to be confused with C. limbatus and C. brevipinna.
The Bull Shark Carcharhinus leucas is one of the more physically distinctive carcharhinid species, due to its heavy-set body and short, blunt snout (Compagno et al., 2005). However, in the Indo-West Pacific its distribution overlaps with that of the Pigeye Shark Carcharhinus amboinensis, a species that shares these same traits to a remarkable degree (Figure 2). Both of these species have been reported from fishery landings in Kuwait, Qatar and the United Arab Emirates (Moore et al., 2012b; Jabado et al., 2014; Bishop et al., 2016), and it seems likely that they occur throughout the Gulf. However, while C. leucas is known to enter freshwater systems including those of Iraq (Ali, 2013; Moore, 2018) and Iran (Coad, 2015), C. amboinensis is not known to share this ability. The occurrence of sharks in freshwater has led to another taxonomic issue, namely the possible presence of the poorly-known Ganges Shark Glyphis gangeticus in the northern Gulf (Compagno, 1984; Moore, 2012a). Khalaf (1961) and Mahdi (1962) have both reported sharks from the Tigris that they identified as G. gangeticus, but it seems likely that these may have been misidentified specimens of C. leucas (Ali, 2013). Reports of sharks from rivers in Iran have been limited to the Whitecheek Shark Carcharhinus dussumieri (in a saline river) and C. leucas (Coad, 2015).
One carcharhinid of particular note is the Smoothtooth Blacktip Shark Carcharhinus leiodon. Prior to 2008, this species was known only from the holotype, which was collected off the coast of Yemen in 1902. Moore et al. (2011) encountered 25 specimens of this species during fish market surveys in Kuwait in 2008, approximately 3000 km from the type locality. The species has since been reported from Salalah in southern Oman (Henderson and Reeve, 2011), which is relatively close to the type locality, as well as the Gulf waters of the UAE (Moore et al., 2013). According to Compagno et al. (2005) C. leiodon could be confused with C. amblyrhynchoides, but all other black-tipped carcharhinids in the region (C. brevipinna, C. limbatus, C. melanopterus) might also pose difficulties for novice data collectors.
Hammerhead Sharks (Sphyrnidae) also pose identification problems in the region. Four species (Winghead Shark Eusphyra blochii, Scalloped Hammerhead Sphyrna lewini, Great Hammerhead Sphyrna mokarran, Smooth Hammerhead Sphyrna zygaena) have distributions that include the northern Indian Ocean and therefore could potentially occur in the Gulf (Compagno et al., 2005). Eusphyra blochii, S. lewini, and S. zygaena have been reported from Iraqi waters, but Ali (2013) suggested that the former and the latter may have been misidentified specimens of S. mokarran and also indicated that S. lewini does not occur in the Gulf. Interestingly, the Smalleye Hammerhead Sphyrna tudes has also been reported from Iraq (Al-Daham, 1974), but this is almost certainly a misidentification given that the species is not known to occur outside the Atlantic Ocean (Compagno et al., 2005; Ali, 2013). Moore et al. (2012b) confirmed the presence of S. mokarran from Kuwait through genetic analysis of study specimens, so there is no doubt that this species occurs in Gulf waters. These authors also noted that although they did not find any evidence of S. lewini or S. zygaena in Gulf fishery landings, a photograph sent to them from Qatar clearly shows a specimen of the former (Moore et al., 2012b). Furthermore, Robinson et al. (2013) reported sightings of S. lewini schools within the Gulf, Jabado et al. (2014) reported S. lewini from fish markets on the Gulf coast of the UAE, and Sabet et al. (2018) recorded S. lewini among trawler bycatch in Iranian Gulf waters, off both Bushehr Province and Hormozgan Province.
Although S. mokarran displays a characteristically straight cephalofoil (Compagno et al., 2005), the difference in cephalofoil morphology between S. lewini and S. zygaena is subtler. Taxonomic keys note that the anterior margin of the cephalofoil in S. lewini exhibits a distinct medial notch, a feature that is lacking in S. zygaena (Compagno, 1984; Compagno et al., 2005; White et al., 2006). However, specimens in fishery landings are not always in pristine condition and the distinction in real life is not always unambiguous (Figure 3).
The last two shark species that present taxonomic challenges are from different genera. The Hooktooth Shark Chaenogaleus macrostoma and the Slender Weasel Shark Paragaleus randalli (both Hemigaleidae) are of similar size (generally <1 m) and are remarkably similar in appearance (Figure 4). Both have been reported from the Gulf (Moore et al., 2012b), and as in the case of the smaller carcharhinid species, they are commonly landed together. Although C. macrostoma is described as possessing characteristically large gill slits and a broader snout than P. randalli (Compagno et al., 2005), considerable familiarity with both species is required in order for routine identification of either species to be made in the field.
The matter of potential cryptic shark species in the Gulf remains to be explored in detail. However, a molecular study by Henderson et al. (2016) on sharks from Omani waters determined that a species occurring in the Gulf of Oman that had long been assumed to be the Bramble Shark Echinorhinus brucus was genetically distinct from this species and from the only other currently recognised echinorhinid species, i.e. the Prickly Shark Echinorhinus cookei (Figure 5). Consequently, these authors suggested that this was a hitherto undescribed species. A formal description of this species has yet to be undertaken, but given the preference of echinorhinid sharks for deeper waters, this particular matter is unlikely to be of relevance to Gulf fisheries. It does, however, highlight the potential for the occurrence of cryptic shark species in the region, and the need to employ molecular phylogeny (as opposed to simple genetic barcoding of individual specimens) in regional species surveys.
One potential case of cryptic speciation that might be relevant to Gulf shark fisheries pertains to L. marcorhinus. Naylor et al. (2012) reported two distinct clusters within their molecular assessment of specimens from throughout the species’ range, while Henderson at al. (2016) found that specimens from Omani waters were similarly clustered. Although there was strong branch support for these clusters, between-group genetic distances were low. Nevertheless, the matter warrants further investigation as some established shark species have been shown to exhibit unusually high genetic similarity with particular congeners, e.g. the Sandbar Shark Carcharhinus plumbeus and the Bignose Shark Carcharhinus altimus (Henderson et al., 2016), and this is likely the case for other shark species as well.
Stingrays, Butterfly Rays and Devilrays
The taxonomic situation regarding rays in the region is particularly challenging. It has long been suspected that certain nominal dasyatid (Dasyatidae) species are species complexes, and recent investigations have confirmed this to be the case (Last et al., 2016a). Perhaps the most problematic of these have been plainly coloured specimens with or without small white spots on the posterior disc and with black and white banding to varying degrees on the tail, occurring within the Gulf and throughout the northern Indian Ocean. Such animals were commonly identified as Himantura gerrardi (now Maculabatis gerrardi), the Whitespotted Whipray. Last et al. (2012) determined that specimens they examined from the Gulf were a distinct species, which they erected as Himantura randalli (now Maculabatis randalli). It seems likely that reports of the Pacific Hemitrygon bennetti from Iranian waters (Paighambari and Daliri, 2012; Kazemi et al., 2014; Sabet et al., 2018) are misidentified M. randalli (Last et al., 2012). Other species within this complex in the northern Indian Ocean have also been resolved in recent years, including Baraka’s Whipray Maculabatis ambigua, the Pakistan Whipray Maculabatis arabica, and Shorttail Whipray Maculabatis bineeshi. None of these species, including M. gerrardi, are currently thought to occur within the Gulf (Last et al., 2016b), but given the complexity of the situation and how little is known about the full distributions of newly established species, considerable care is recommended when identifying Maculabatis specimens in Gulf fishery landings. Three ray specimens reported as Himantura sp. from Fujairah on the Gulf of Oman coast of the UAE by Henderson et al. (2016) are now known to be M. arabica, based on NADH2 sequences (G. J. P. Naylor, University of Florida, Gainesville, FL, USA, pers. comm.). This extends the known range of this species to within ∼150 km of the Strait of Hormuz, so it seems plausible that the species may also occur within the Gulf.
Confusion between the Leopard Whipray Himantura leoparda and the Coach Whipray Himantura uarnak is also a cause for concern. Adults of the former exhibit a “dense pattern of dark medium-size rings” while those of the latter are “densely covered with very small brownish black spots and flecks” (Last et al., 2016b), and they are therefore reasonably easy to distinguish from each other. However, the two species display ontogenetic changes in colour patterning, and both go through phases of exhibiting solid black spots (Last et al., 2016b), during which they are considerably more difficult to tell apart.
Until relatively recently, the genus Pastinachus was thought to contain a single species, i.e. the Cowtail Ray Pastinachus sephen, with a widespread Indo-Pacific distribution (Randall, 1995). Taxonomic investigations have resulted in the establishment of five additional species within the genus, and P. sephen is now thought to be restricted to the northwestern Indian Ocean, including the Gulf, where it is sympatric with the Broad Cowtail Ray Pastinachus ater (Last et al., 2016b). According to Last et al. (2016b) these two species differ only subtly in morphology and denticle structure, hence they are extremely difficult to separate from each other in the field. Four Pastinachus specimens from the Gulf of Oman and Arabian Sea encountered by Henderson et al. (2016) that were tentatively identified as P. sephen (Figure 6), were later confirmed to be P. ater after assessing their NADH2 sequences.
There is some doubt surrounding the true identity of the Longtail Butterfly Ray Gymnura poecilura in the region. In their global-scale molecular assessment of elasmobranch taxonomy, Naylor et al. (2012) noted that specimens identified as G. poecilura formed two distinct clusters which they designated as G. cf. poecilura 1 and G. cf. poecilura 2. It has since emerged that the former is likely to be the true G. poecilura while the latter is possibly an undescribed species (G. J. P. Naylor, University of Florida, Gainesville, FL, USA, pers. comm.), a conclusion supported by molecular comparison of specimens from the Bay of Bengal and the northern Arabian Sea (Muktha et al., 2016). All of the ‘G. poecilura’ specimens assessed by Henderson et al. (2016), which included representatives from the Gulf, formed three distinct sub-clusters within G. cf. poecilura 2. As in the case of L. macrorhinus (above), there was strong branch support for these sub-clusters, but genetic distances between them were low, so it is unclear if they represent cryptic lineage divergence. Regardless of whether these specimens represent one or multiple species, it seems likely that Longtail Butterfly Rays reported from the Gulf are something other than G. poecilura.
Lastly, a recent molecular assessment of the Longhorn Devilray Mobula eregoodootenkee, which has been reported from the Gulf (Di Sciara et al., 2017), suggested that it is conspecific with the sympatric Kuhl’s Devilray Mobula kuhlii (White et al., 2017). However, subsequent molecular work has revealed that the two may indeed be separate species (Hosegood et al. 2018). Therefore, it would seem prudent for data collectors to continue treating these Devilrays as separate species based on their distinguishing character (relative cephalic lobe length), until this particular taxonomic matter is fully resolved.
Guitarfishes and Wedgefishes
Last et al. (2016b) list three guitarfish species (the Spotted Guitarfish, Rhinobatos punctifer, Sharpnose Guitarfish Glaucostegus granulatus, Halavi Guitarfish Glaucostegus halavi) and four wedgefish species (the Bowmouth Guitarfish or Shark Ray Rhina ancylostoma, Bottlenose Wedgefish Rhynchobatus australiae, Whitespotted Wedgefish Rhynchobatus djiddensis, Smoothnose Wedgefish Rhynchobatus laevis) as occurring in the Gulf, but there is still some confusion regarding the identities and distributions of Guitarfishes and Wedgefishes throughout the region. For example, the Bengal Guitarfish Rhinobatos annandalei (which is known to occur in the Gulf of Oman and Arabian Sea) and the Bottlenose Guitarfish Rhinobatos schlegelii (which is only known from the northwestern Pacific) have been reported from northern Gulf waters (Ali, 2013; Kazemi et al., 2014). Records such as these are most likely due to confusion over the physical appearance of R. punctifer.
Rhinobatos punctifer is somewhat unusual in that it has been formally described and rediscribed the same authors (Compagno and Randall, 1987; Randall and Compagno, 1995). A photograph of the specimen from the latter publication was subsequently printed in colour in Randall (1995), one of the most comprehensive sources of taxonomic information on the region’s ichthyofauna, and the basis for the identification of R. punctifer specimens from Omani waters by Henderson et al. (2007; 2016). Guitarfish with similar colouration but a notably narrower disc are also commonly encountered in Omani fishery landings and such specimens were treated as a potentially separate species by Henderson et al. (2007), who assigned the three-letter species code ‘RHY’ to them (‘RHP’ being the designation for specimens identified as R. punctifer). Analysis of NADH2 sequences subsequently confirmed that RHP and RHY specimens are distinct species, with the latter assumed to be an unidentified Rhinobatos sp. (Henderson et al., 2016). However, closer inspection of the R. punctifer specimens presented in the aforementioned descriptions (Compagno and Randall, 1987; Randall and Compagno, 1995) indicates that they might actually be two different species. The R. punctifer specimen illustrated in Randall and Compagno (1995) and Randall (1995) seems to more closely resemble R. annandalei, as described by Last et al. (2016b). Consequently, any guitarfish that have been identified as R. punctifer based on the descriptions in Randall and Compagno (1995) and Randall (1995) are most probably R. annandalei, e.g. Henderson et al. (2007; 2016), Henderson and Reeve (2011). Conversely, the RHY (Rhinobatos sp.) specimens reported by Henderson et al. (2007; 2016) are most likely R. punctifer and not an undescribed species (Figure 7).
The situation is further complicated by the fact that the dorsal colour pattern on R. punctifer appears to be highly variable. The Rhinobatos sp. reported by Henderson et al. (2016), and which is now thought to be R. punctifer (see above), was noted to commonly exhibit ocellated spots and sometimes more elaborate patterning (Figure 8). The extent of this patterning was such that Henderson et al. (2016) initially thought that these patterned individuals might represent another, distinct species, but analysis of NADH2 sequences revealed them to be conspecific. Consequently, it is important to note that R. punctifer specimens may exhibit solid spots, ocellated spots, or a combination of large ocellated spots and prominent reticular patterning. Rhinobatos annandalei, on the other hand, always exhibits solid spots. Further clarification on the matter will be provided in a forthcoming publication (P. R. Last, CSIRO, Tasmania, Australia, pers. comm.).
Until relatively recently, only two Wedgefish species were believed to occur in the Gulf and adjacent waters, i.e. Rhina ancylostoma and Rhynchobatus djiddensis, which are physically quite distinct from each other. However, Henderson et al. (2016) noted that specimens of R. djiddensis from the region formed two very distinct clusters based on NADH2 sequences, which they designated R. cf. djiddensis 1 and R. cf. djiddensis 2. Subsequent taxonomic assessment of the Rhinidae has determined that R. cf. djiddensis 1 is actually R. laevis while R. cf. djiddensis 2 is R. djiddensis (G. J. P. Naylor, University of Florida, Gainesville, FL, USA, pers. comm.), and specimens of both were collected from Gulf fishery landings (Henderson et al., 2016). As mentioned previously, Last et al. (2016b) report that R. australiae also occurs in the Gulf and adjacent seas, but it is not clear how this conclusion was reached. What is clear though, is there is remarkable physical variation in Rhynchobatus specimens encountered in regional fishery landings (Moore, 2012b; Henderson et al., 2016), and further research will almost certainly yield additional species records and range extensions.
The implementation of effective management measures is urgently required in order to ensure the conservation of Gulf elasmobranch stocks (Jabado and Spaet, 2017). Some taxa, such as the sawfishes (Pristidae), were once common in the region but are now functionally extinct (Moore, 2015), while landings of others continue to decline despite increasing fishing effort (Jabado, 2018). Numerous matters need to be addressed in order to pave the way for meaningful management measures in the region (Moore, 2012a), but as mentioned in the introduction to this article, the fundamental requirement in species-specific data collection is that species are correctly identified. Field staff concerned with elasmobranch data collection in Gulf fisheries should treat the information contained herein as a precautionary warning regarding species that may be difficult to definitively identify in a field setting, even for experienced researchers, but it is important to note that this paper does not present an exhaustive account of the elasmobranch species occurring in Gulf waters or fishery landings. The reality of the matter is that all species are open to potential misidentification, and a great deal depends on the quality of training and resources provided to field staff and the diligence with which they execute their duties.
I am extremely grateful to A. Moore, G. Naylor and P. Last for their comments surrounding elasmobranch taxonomic issues in the region.
Aaron C. Henderson http://orcid.org/0000-0001-8873-0785
This article has been republished with minor changes. These changes do not impact the academic content of the article.