Back from obscurity : notes on the current distribution , threats and conservation status of a poorly known cyprinid , Hypselobarbus lithopidos ( Day , 1874 ) from the Western Ghats of India

Author Details: Anvar Ali interested in taxonomy and systematics of freshwater fishes of the Western Ghats. Siby Philip is interested in molecular phylogenetics, evolution and biogeography of freshwater fishes of the South Asia region. Rajeev Raghavan is interested in interdisciplinary research focused on generating information and developing methods to support conservation decision-making, especially in freshwater ecosystems


INTRODUCTION
In his paper on 'Some new or little known fishes of India' published in the Proceedings of the Zoological Society of London, Francis Day (1874) described five new species of fish, one of which was Barbus (= Hypselobarbus) lithopidos, from South Canara (= Dakshina Kannada).The fish was known to attain up to 24 inches in length and was considered 'not uncommon' in rivers of the region (Day 1878(Day , 1889)).A specimen measuring 7.4 inches long was illustrated by Day (1878) in his book on the Fishes of India; being a natural history of the fishes known to inhabit the seas and fresh waters of India, Burma, andCeylon. From 1909 until 1975, several collections of H. lithopidos were made from the rivers of the southern part of the Western Ghats flowing through the states of Kerala and Karnataka (see table 1), after which there were no records of the species.The absence of H. lithopidos in collections from a region which has been the focus of several comprehensive field surveys (Easa & Basha 1995;Chapgar & Mankadan 2000;Kurup et al. 2004;Thomas 2004;Biju 2005) led to the assumption that the species might probably be extinct (see Raghavan & Ali 2011).
Although several recent papers (Arunachalam et al. 2000;Vijaykumar et al. 2008;Ahmad et al. 2013;Vijaylaxmi & Vijaykumar 2011;Arunachalam et al. 2012) mentioned collecting H. lithopidos, these need to be verified as no voucher specimens are available.Ahmad et al. (2013)  not provide this record, probably due to the lack of clarity in the species identification.There was also a general consensus within the working group that there are no validated records of this species since 1941 (see Raghavan & Ali 2011).
Similarly, Arunachalam et al. (2012;pp. 70 and 71) in their revision of the genus Hypselobarbus records collecting H. lithopidos from the Shimoga fish farm and Rusewalai (= Rosemalai?)fish farm.They also provide illustrations for these materials (see figure 1A p 64 and figure 3A p 66 of Arunachalam et al. 2012).However, these are clear misidentifications as the material illustrated in p 64 has a lateral line scale count of 30-31, while the material in p 66 has a lateral line scale count of 34-36.Hypselobarbus lithopidos has a lateral line scale count of 37-39 (Day 1874;Knight et al. 2013; also see Table 2).Raghavan & Ali (2011) also suggested that the record of H. lithopidos by Arunachalam et al. (2000) from Kal River in the Konkan region of Maharashtra requires verification.This suggestion might hold true as the record of H. lithopidos from Kal River, did not find a mention in a recent paper by the same authors (Arunachalam et al. 2012).
Recently (Knight et al. 2013) re-discovered H. lithopidos based on collections from the Phalguni River in Dakshin Kannada District of Karnataka State, laying to rest the speculation that the species is extinct.As part of the on-going project on the 'lost fishes of Western Ghats', we obtained specimens of H. lithopidos (Image 1) from an additional river system, part of the erstwhile South Canara region, its type locality.This paper serves to document this additional record, and discuss the current distribution, threats and conservation status of this species.We also propose a revised Red List status for H. lithopidos.Francis Day, BMNH 1889.2.1.559-561, 3 exs., 124.72-171.42mm SL, South Canara, India (Image 2); Hypselobarbus lithopidos, CRG-SAC.2013.3.1-5, 10.iii.2013; 5 exs., 49.63-60.93mm SL, Pallamkode, 12.564 0 N & 75.246 0 E, Chandragiri River system, Kerala, India, Anvar Ali.
Taxonomy: We use the generic name Hypselobarbus instead of Gonoproktopterus following Arunachalam et al. (2012) and Yang et al. (2012).
Morphometric data collection: Counts and measurements follow Pethiyagoda et al. (2012).Measurements were taken using a digital calliper to the nearest 0.1mm.Subunits of body are presented as percent of standard length (SL) and subunits of head are provided as percent of head length (HL).

RESULTS AND DISCUSSION
Morphometric characters and meristics are presented in Table 2 and Appendix 1.For detailed description refer to Knight et al. (2013).Values of morphometric characters of the materials that we collected (Table 2) show significant variations from those recorded by Knight et al. (2013).Our specimens were juveniles (49-60 mm), whereas those collected by Knight et al. (2013) and Day (1874) were adults.Raj (1941) mentions that young specimens of H. lithopidos (57-90mm) show variations in the morphometric characters and coloration when compared to adults.In addition, it is known that closely related large growing cypriniform fishes, often tend to have an allometric growth pattern (Mina et al. 2005;Patimar & Farzi 2011), which explains the discrepancy between the morphometric proportions in our study.
Except for the pre-dorsal scales, the meristics of the materials collected by us match those of the syntypes at the Natural History Museum (BMNH), London and those recorded recently by Knight et al. (2013).The pre-dorsal scale count of our materials was 12-13, while those of Knight et al. (2013) and the syntypes/Day's materials at BMNH and AMS are in the range of 13-14 (R. Britz & R. Pethiyagoda pers.comm).Raj (1941), mentioned that H. lithopidos has a pre-dorsal scale count range of 11-14 based on the observations of both his material (from Nilambur) as well as Day's materials in the Zoological Survey of India, thereby bringing the pre-dorsal scale count of the specimens examined in our study well within range.
Interestingly, not a single individual of H. lithopidos has been recorded from the state of Kerala in the last four decades.The last individual recorded from this part of the Western Ghats was from the Periyar Tiger Reserve in 1975 by R.S. Pillai (Jayaram 1991).Since then, several comprehensive ichthyological surveys have been carried out in the Periyar Tiger Reserve resulting in the description of many new species, but no records of H. lithopidos have been made.Although voucher specimens of R.S. Pillai's collection (ZSI.SRS.F.2088; 3 ex.148-158mm) were deposited, these are currently untraceable (see Knight et al. 2013).Also, previous records of H. lithopidos from Kerala with the exception of Raj (1941) do not provide detailed information (especially biometrics) on the materials collected or examined.Raj (1941) recorded a single specimen from Nilambur for  2012), ichthyological books/field guides from India (for e.g.Tekriwal & Rao 1999) have also provided misleading photographs on the identity of H. lithopidos.Tekriwal & Rao (1999;p 65) in their book on the ornamental aquarium fish of India provide an image of a barb that they call Gonoproktopterus (= Hypselobarbus) lithopidos.However, the fish in the image has no more than 25 lateral line scales, and therefore is a misidentification.We hope that the images of live specimens as well as those of the type series of H. lithopidos provided in Knight et al. (2013) and the current paper will help ichthyologists working in the Western Ghats region to correctly identify H. lithopidos in the future.
The information available at present indicates that H. lithopidos occurs in the west flowing Phalguni stream (Kumaradhara-Nethravati river system) and the Chandragiri River in the state of Kerala (Image 4).Both these rivers are part of the erstwhile South Canara from where the species was described by Day (1874).The records further south (earlier records; Image 4) could either be misidentifications, or point to a scenario where the species was probably extirpated from these locations.However, both these scenarios need to be confirmed with future exploratory surveys.
The habitat in Pallamkode, from where we collected specimens of H. lithopidos, consist of many deep pools with intermittent 'run' and 'glide' microhabitats, and supported by high amount of canopy cover.In-stream habitat cover existed in the form of twigs and wooden logs.
Local knowledge of fishers in the Chandragiri River indicate that the species is mostly found in the streams from Sullia in Karnataka to (near) the town of Adoor across the border in Kerala.Fingerlings are especially seen during the months of July to August which might indicate that the species breeds during the south-west monsoons.Our collection of juveniles also indicates that recruitment is taking place, and if on-going threats are curtailed populations of this rare large barb could be stabilized.
Interestingly, extensive surveys in the Kasargod District of Kerala (erstwhile South Canara) including River Chandragiri by Biju (2005) as well as neighbouring basins of Uppala and Manjeshwaram by Biju et al. (1999 a,b) did not yield any specimens of H. lithopidos.These authors could also not collect H. thomassi, known to be the closest resembling congener to H. lithopidos (see Knight et al. 2013), thereby eliminating the possibility of a misidentification between the two.Dynamiting is a major threat to the species in the streams around Sullia (Karnataka State), where there are many deep pools in which large sized H. lithopidos are often seen.Poisoning and electric fishing are also prevalent in this region, especially when the water levels are low.There is also a continuing decline in the quality of habitats due to pollution from domestic sources as well as petroleum related pollution as a result of washing motor vehicles.
There is no targeted fishing for H. lithopidos in the Chandragiri River Basin.However, they are often caught as by-catch during aquarium fish collections for two congeneric species, H. jerdoni and H. thomassi.Local aquarium fish collectors reveal that the species is often mixed with consignments of H. jerdoni as a filler, since both species look superficially alike in their juvenile stages.
Currently, H. lithopidos has been listed as Data Deficient (Raghavan & Ali 2011) with a note that the species could probably be extinct.Additional information from Knight et al. (2013) and the present paper have led to a scenario where the Red List status of the species needs to be revised.The proposed Red List Status of the species has been provided in Appendix 2. Rediscovery of purportedly extinct and missing species is known to be a non-random process, with the chance of success depending on search effort, search area, time missing and traits known to be associated with extinction risk such as population density and range size (Fisher & Blomberg 2011).Considering the increasing spate of rediscoveries across several taxa, Scheffers et al. (2011) believe that many species, particularly those known only from type specimens, that are currently thought to have gone 'extinct', remain 'extant', and with continued support for biological surveys, many of these species will be relocated with time.Several recent studies on freshwater fishes of the Western Ghats which have reported the rediscovery of 'missing', 'long lost' or 'poorly known' species (Dahanukar et al. 2011;Britz et al. 2012;Ali et al. 2013;Knight et al. 2013) listed H. lithopidos as occurring in Hariharapura in the Tunga drainage.However, one of the authors of the paper who was part of the Cypriniformes working group that evaluated the conservation status of the Hypselobarbus species from Western Ghats in 2010 for the IUCN Red List of Threatened Species did Image 1. Hypselobarbus lithopidos CRG-SAC.2013.3collected from Chandragiri River, Kerala © Rajeev Raghavan

Image 4 .
Current and previously recorded range of Hypselobarbus lithopidos in the Western Ghats

Table 2 . Morphometric characters of Hyselobarbus lithopidos collected from Chandragiri River, Kerala Morphometric character Range (n=5) Mean+SD
are testimony to this.