First pelagic record of the velvet dogfish Zameus squamulosus (Günther, 1877) (Squaliformes) from the southwestern Indian Ocean and some notes on its regional distribution

ABSTRACT A pelagic record of a rare deep-water shark, the velvet dogfish Zameus squamulosus (Günther, 1877), is described from the southwestern Indian Ocean. This is the first pelagic record from the western Indian Ocean and the eleventh published record of this species from the entire basin. Together with non-published records from museums and online databases the number of verified Indian Ocean records of this species currently exceeds 50 individuals. Zameus squamulosus is a benthopelagic species usually occurring on the slopes of the continents and in mid-ocean oceanic ridges, between 400 and 1450 m depth, but it makes rare incursions in open water to the limits of the epipelagic zone.


INTRODUCTION
The squaloid shark Zameus squamulosus (Günther, 1877) is a deep-water species, encountered in the benthic and sometimes in the pelagic environment along continental margins of tropical and temperate waters in the Atlantic and Pacific (Compagno 1984;Wetherbee & Crow 1996). Its distribution and occurrence in the Indian Ocean are still obscure. Only few published records of Z. squamulosus are known from the undersea ridges of the southern tropical Indian Ocean (Scherbachev et al. 1982). Other records, although lacking geo-referencing precision, are attributed to continental slope areas of South Africa, Australia and Java (Indonesia) (Bass et al. 1976;Last & Stevens 1994, 2009White et al. 2006). One individual was recently collected at ABSTRACT A pelagic record of a rare deep-water shark, the velvet dogfish Zameus squamulosus (Günther, 1877), is described from the southwestern Indian Ocean. This is the first pelagic record from the western Indian Ocean and the eleventh published record of this species from the entire basin. Together with non-published records from museums and online databases the number of verified Indian Ocean records of this species currently exceeds 50 individuals. Zameus squamulosus is a benthopelagic species usually occurring on the slopes of the continents and in mid-ocean oceanic ridges, between 400 and 1450 m depth, but it makes rare incursions in open water to the limits of the epipelagic zone.
Despite its global distribution, biology and habitat of the velvet dogfish are still poorly known: it is listed as "data deficient" on the IUCN Red List of Threatened Species (Burgess & Chin 2006). Therefore information presented in this note is important for the understanding of this species' ecology.
Here we describe the first record of Z. squamulosus from the pelagic zone of the tropical southwestern Indian Ocean and present a further nine recent records from the demersal environment. We discuss regional species distribution based on: 1) published data; 2) museum collection inventories; and 3) online sources. Finally, we debate the reliability of information from internet portals.   Table 2 for details) (ruler length is 30 cm).
Romanov E. V. et al.
Another eight specimens of this species were caught by demersal longlines during commercial fishing operations in the southern Indian Ocean in March-April 2008. Individuals were sampled and identified by S.T. Rebik as Scymnodon obscurus Vaillant, 1888 (a synonym of Zameus squamulosus) using keys in Compagno (1984) (Fig. 1B). Denticles from left side of the trunk were also photographed using a light microscope (Fig. 3B). Occurrence, fishing depth and some biological observations were recorded and used here to illustrate both spatial and vertical species distribution (Figs 2; 4).
One more individual (female, 710 mm TL) preserved in the MNHN collections (catalogue number MNHN 2007-1658, identified by B. Séret as S. obscurus) was caught on the slope of Reunion Island using a demersal longline baited with squid on 22 February 2005 by the commercial fishing vessel Ludo (Fig. 2).
The Indian Ocean area is considered here according to FAO definitions of fishing areas 51 and 57, with its western border off South Africa at 30°E and eastern border off Southern Australia at 150°E. data taken from the GEBCO database (GEBCO 2010) suggests that the fishing gear was deployed and drifted over a deep-water canyon (bottom depths 5000-5400 m) surrounded by deep-water seamounts (minimum depth 3500-3800 m), and by an abyssal plain of 4500-4900 m deep.
The shark was frozen and stored for 30 days before examination. The specimen was deposited in the collections of the Muséum national d'Histoire naturelle, Paris (catalogue number MNHN 2012-0188).
Measurements were taken to the nearest mm following Bass et al. (1976) and Compagno (1984) approaches and expressed as percentage of TL. Total length used for calculation of body proportions was measured for "stretched" caudal fin as presented in Compagno (1984) in alternative to "natural" position as suggested by Bass et al. (1976). Denticles from left side of the trunk (Fig. 3A) were examined under a light microscope (Olympus SZ61 under ×45 magnification) and photographed with digital camera Olympus SP510UZ resolution 7.1 Mpx.  (Table 2): Pelagic records: yellow dot (this study) and orange dots (CSIRO); Demersal records: black dots (YugNIRO, this study), grey dots ZM MGU (Scherbachev et al. 1982

OBIS
The    (2001), except for terms and measurements adopted from Bass et al. (1976) given in italics, and spine measurements adopted from Clarke, Irvine (2006), given in bold italics.

ZOOSYSTEMA • 2013 • 35 (1)
Our specimen well corresponds to the original description given by Günther (1877) and later by Yano & Tanaka (1984). Body measurements are within the ranges presented in other studies, except PP2, SVL (slightly higher than reported earlier) and trunk width (slightly lower) ( Table 1). The latter could be explained by post-defrosting deformation.

DISCUSSION
To date, 10 individuals of this species are reported in the literature from the Indian Ocean region (Table 2). First record by Bass et al. (1976) who reported four semi-digested individuals taken from the stomach of a sperm whale Physeter macrocephalus harpooned 112 km southeast of Durban (Fig. 2). Consequently no exact data on both shark location and habitat depth were recovered. Another five individuals were caught by bottom trawls along deep-water ridges of the southern tropical Indian Ocean during Soviet research cruises in 1976-79 (Scherbachev et al. 1982). One more individual was recently recovered at the landing site in Cochin Fishing Harbour (India), but precise capture position is apparently unknown (Akhilesh et al. 2011) ( Fig. 2; Table 2).
Other combinations: Scymnodon squamulosus (Regan, 1908), Scymnodon obscurus Bigelow and Schroeder (1957 comm., 2011). In addition, capture positions provided by GBIF and OBIS for the same individuals (inventory numbers) from the collections of the Australian Museum and the Museum Victoria were different. Therefore, the reliability of online data sources for reporting of rare species distribution is still questionable as suggested by Eschmeyer & Fricke (2012).
Our records and earlier collected data show that Z. squamulosus is distributed within the subtropical area of the southern Indian Ocean from the African to the Australian coasts (Fig. 2). Only few records are from the tropics and only one individual is from the northern hemisphere (Fig. 2). Capture positions are usually associated with continental slopes and mid-ocean undersea ridges.
The majority of Indian Ocean records except our pelagic individual and twelve individuals from CSIRO were caught in demersal environment. Similarly White et al. (2006) reported occurrences of this species in Indonesia (off western Java).
Direct correspondence with museums allowed us to obtain data on 35 records from the Indian Ocean (CSIRO, 31 ind.; Iziko SAM, 3 ind.; WAM, 1 ind.). Among these records twelve originated from pelagic gears, 22 were caught in the demersal environment, and one (WAM) was found in the stomach of a sperm whale ( Fig. 2; Table 2).
Online resources were assumed to be a handy tool for fast and easy access to taxonomic and biogeographic information (Costello & Berghe 2006). Indeed internet portals appeared to be very useful to find occurrences, which otherwise would remains unknown/inaccessible for us. However we found that online portals handling taxonomic and fish biodiversity information (Discover Life, Fish-Base, GBIF, ITIS, OBIS, WoRMS) often maintain outdated taxonomic data. The species in question was commonly found either under synonymised names as S. squamulosus and S. obscurus or both were considered as separate valid species (  The range of the TL of the Indian Ocean individuals reported in this note is within 415 and 900 mm. Four of eight individuals analysed in the field by STR exceeded the maximum reported length (TL) of 786 mm (Wetherbee & Crow 1996), and one exceeds the maximum length (TL) suggested by Last & Stevens (1994, 2009 (Tables 1; 2). Both sexes were present in catches; some individuals were mature. One gravid female (TL = 900 mm) caught on 17 March 2008 (Table 2) had two embryos at an advanced stage of development. Overall Indian Ocean records show dominance of females: 1.9:1.0, which however non-significantly deviate from 1:1 sex ratio (χ 2 = 3.125, p = 0.0771).
The rarity of this species is an apparent consequence of a relatively low fishing effort within its principal depth distribution range and the scarcity of scientific observations in deep-water fisheries. However low fishing pressure is rather beneficial for the conservation of this species taking into account both the high longevity and slow reproduction rate characteristic of most deep-water chondrichthyans (Musick & Bonfil 2005). crew of F/V Antillas Reefer for their help rendered to STR during sampling and photographing. Our thanks to Mr Lausin Beboit, captain of F/V Ludo, for the collection and preservation of an individual of Z. squamulosus (MNHN 2007(MNHN -1658.
Contributions of referees: S. Amir, P. Kyne and A. Ohler improved clarity and completeness of the paper.
Special thanks are due to Clément Martineau (ESA, Angers) for his help in microscopic photography of scales.
Alastair Graham (CSIRO), Sue Morrison (WAM), Michael Bougaardt (Iziko SAM), and Roger Bills (SAIAB, Grahamstown) provided data for specimens 1400 m depth; peak of occurrences is within the 1000-1200 m range (Fig. 4). Such a pattern suggests a benthopelagic life style of Z. squamulosus with regular excursions from the benthic to pelagic waters within the upper bathyal habitat.
Meanwhile, our pelagic capture reported here is not an exception. In the Atlantic and Pacific Ocean Z. squamulosus is considered as a pelagic species in some areas (Yano & Tanaka 1984;Compagno 1984). Numbers of pelagic records of this species have increased exponentially worldwide from single individuals some decades ago (Krefft 1980;Wetherbee & Crow 1996) to several dozens of individuals reported annually in recent years (Matsuhita & Matsunaga 2002;Dai et al. 2009;Zhu et al. 2012). Last & Stevens (2009) stated that pelagic records are more common than benthic ones off southern Australia. Such an increased number of recent occurrences cannot be explained only by expanded scientific observers' coverage of fishing operations and improved identification of specimens caught. The archive of high-quality long-term research data collected with pelagic longlines over all the Indian Ocean (Romanov et al. 2006) contains no records of this species in the pelagic environment from 1961 to 1989. We may be evidencing a long-term increase in abundance of this species similarly with other mid-sized oceanic predators like crocodile shark (Romanov et al. 2008), bramids and gempylids (Polovina et al. 2009) in response to the depleted state of large-sized predators or competitors (mesopredator-release effect) (Baum & Worm 2009;Ferretti et al. 2010).
At the same time, both pelagic and benthic records of Z. squamulosus are still rare over the vast area of the central and western Indian Ocean. Pelagic longline fisheries covering most of the surface of the tropical and temperate Indian Ocean developed an average fishing effort of almost 700 million hooks annually from 2000 to 2010 (Fonteneau 2010; Fonteneau pers. comm., 2012). Only thirteen pelagic individuals of this species recorded during the history of regional fisheries observations under such significant fishing effort suggests that epipelagic excursions should be considered rather as an exception than common habit.