Distribution Pattern of Seahorse Species (Genus: Hippocampus) in Tamilnadu and Kerala Coasts of India

The survey along the Tamilnadu and Kerala coats of India reveled that six species of seahors (Hippocampus fuscus, H. kelloggi, H. kuda, H. histrix, H. mohnikei and H. trimaculatus) were distributed with different density. Out of the six species, H. fuscus, H. kuda and H. trimaculatus, were the commonly available in all the observed areas. In Palk Bay, H. kuda was the dominant species constituting 49.10% of the total seahorses encountered. Hippocampus trimaculatus was the second dominant species which accounting 39.28%. The Gulf of Mannar region also most abounded with H. kuda (68.98%) followed by H. trimaculatus (20.80%), H. fuscus (9.80%), H. kelloggi (2.23%) and H. histrix (0.37%). In Kerala coast, H. trimaculatus was the dominant species (79.68%) followed by H. kuda (9.89%), H. kelloggi (8.33%) and H. fuscus (2.08%). To infer the variation of six seahorse species, the morphometric and meristic characters were analysed. The important morphometric and meristic characters are trunk rings, tail rings, pectoral and dorsal fin rays, trunk length, tail length, coronet height, head length, snout length, snout depth and head depth. Variations in overall body shape, relative snout length, coronet height, number of tail ring was sufficient to separate the specimens to Hippocampus fuscus, H. kelloggi, H. kuda, H. histrix, H. mohnikei and H. trimaculatus. The species density and diversity depends on the habitat and biogeography of those areas. Majority of seahorse fishing in Tamilnadu was by shrimp trawl, by-catch and very few target catch by divers also seen in some villages in Palk Bay and Gulf of Mannar region. The shrimp trawl by-catch only bringing more H. trimaculatus than the other seahorse species in Kerala coasts.


Introduction
Seahorses are one of the members of the family Syngnathidae and thirty three species come under the single genera, Hippocampus (Dawson, 1980;Lourie et al., 2004). They are found throughout the world in shallow, coastal tropical and temperate waters and are more abundant in the Indo-Pacific region (Lourie et al., 1999). They are distributed in high density along the China coast among seaweed and in seagrass from Vietnam to Korea (Lourie et al., 1999). Based on the tentative assessment, Vincent (1996) observed that the seas adjacent to China might contain about seven species, although their taxonomy remained confused. In north-east countries like Philippines and Vietnam they often found in seagrass beds, mangrove, coral reef areas and also in estuaries. Among all seahorses, H. trimaculatus was the most common species on sale in Vietnam markets (Vincent, 1996).
The unclear taxonomy of seahorse is due to the limited morphological variation among species, poor type descriptions, independent designation of the same name for different species, and the ability to changing their body colour and growing skin filaments to mach their surroundings (Lourie et al., 1999). As per Lourie et al. (1999), the standard morphometric and meristic character analysis may clear the taxonomical ambiguity. They are: trunk rings, tail rings, pectoral and dorsal fin rays, trunk length, tail length, coronet height, head length, snout length, snout depth and head depth.
Seahorses have been traded along with pipefishes as dried form to East Asian countries for medicinal purpose (Martin-Smith et al., 2003). Now, seahorse fishing and trading is restricted in all over the world because of their declining population size and the entire seahorse species are listed in scheduled (Baillie and Groombridge, 1996). In India the Ministry of Environment and Forests banned the export permits for all Syngnathids from the year 2001 and declared them under Schedule I of the Indian Wildlife Protection Act, 1972 (Sreepada et al., 2002). In India, Marichamy et al. (1993) tentatively reported that two species (H. kuda and H. trimaculatus) was found along the Palk Bay. Later on, Lipton and Thangaraj (2002) made a detailed study of these species and one more species was added (H. fuscus) in Palk Bay coastal region. More recent-tutoyo, Japan) with 0.1 mm accuracy. The meristic counts were counted under a dissection microscope and the values were confirmed by three time counting. All the morphometric measurements and counted characters were strictly followed Lourie et al. (1999) protocol.

Morphological characters and species identification
During the study period, six species of seahorses ( Fig.  1) were recorded from the observation centers. Variation in overall body shape, relative snout length, coronet height, number of tail ring was sufficient to separate the specimens to Hippocampus fuscus, H. kelloggi, H. kuda, H. histrix, H. mohnikei and H. trimaculatus (Fig. 1). The basic morphometric and meristic statistics are given in Tab. 2, the abbreviations are given in Lourie et al. (1999).

Percentage distribution
The distribution of the above six species was not uniform in all the sites surveyed. The percentages of species availability in three locations are given in Tab. 3. Only one specimen of H. mohnikei was observed during this study ly, Murugan et al. (2008) reported five species of seahorses along the South east coast of India. In the present paper, it is described the morphometric and meristic characters of seahorses, and the survey shows the distribution pattern along the Tamilnadu and Kerala states of India.

Seahorse survey
For the seahorse resource assessment studies, Palk Bay, Gulf of Mannar in Tamilnadu and Malabar coast in Kerala was selected and the details are given in Tab. 1. Generally seahorses are caught as by-catch by small trawlers, targeting the shrimp in both the states. The data such as the number, types of boats and duration of operation were collected by interviewing the respective area fisherfolk. Shore seines are used in the shallow coastal region at 2-3 m depth. Country trawl nets are operated at 3-6 m depth. The shrimp trawlers are operated at 15-50 m depth and targeting shrimps. Data collections were made from all the three regions on monthly basis over a period of four years from June 2003 to May 2007. The by-catch of seahorses from all the boats in each area was counted and additional data also gathered from the primary traders in the respective areas.

Morphological characters and species identification
All the wet specimens were obtained as by-catch by small off shore trawlers, targeting the shrimp and the dried specimens were got from small traders in the repective stations. No seahorse was caught especially for this work. With the reference of the identification book (Lourie et al., 1999)

Discussion
Detailed morphometric analysis was sufficient to distinguish this four seahorse morphotypes, although intergroup distance was sometimes small. In this article we have highlighted the characters which permit species identification by eye and this may facilitate practical application in this field. Hippocampus fuscus could be distinguished from the other two related Indian seahorses, viz., H. borboriensis and H. kuda. The morphometric characteristics of H. borboniensis were studied earlier by Salin et al. (2005). Hippocampus borboniensis possesses a relatively larger head and two prominent eye spines on the head. In H. fuscus there is no eye spine. Hippocampus borboniensis has a shallower body than that of H. fuscus. Hippocampus kuda also differs from H. fuscus by having larger body size and backward turned coronet. In H. fuscus, the coronet is slightly raised without any backward curve. In H. fuscus the snout depth also differs from that of H. kuda. In H. fuscus, the snout depth is somewhat higher (30.22 ± 7.15% of snout length) than in H. kuda (28.91 ± 4.01% of the snout length). When the dorsal finbase length (DL) and pectoral finbase length (PL) were compared, H. fuscus has been found to be lower (23.74 ± 4.36% of trunk length) than H. kuda DL (27.13 ± 5.03% of trunk length). In H. fuscus, the PL is smaller (33.38 ± 8.84% of head depth) than that of H. kuda (34.07 ± 7.20% of head depth). In H. fuscus, the head depth is higher (66.05 ± 8.33% of head length) than H. kuda (59. 50 ± 7.50% of head length). Though H. hippocampus is considered to be similar to H. fuscus, it is restricted only to the European waters.
Hippocampus kelloggi is the largest seahorse species found in the Indian waters. Fish Base (Froese and Pauly, 2012) considered the distribution of H. kelloggi as of Indo-Pacific and listed the same as a possible Indian species. Hippocampus kelloggi can be distinctly differs from other morphologically similar species such as H. borboniensis, H. fuscus, H. ingens, H. kuda and H. spinosissimus. Hippocampus kelloggi has five elevated short coronet spines and a plate in front of the coronet, but in the case of H. borboniensis, the coronet has five rounded knobs and no plate in front of the coronet as per Lourie et al. (1999). Tail rings are also higher (39-41) in H. kelloggi than the H. borboniensis (34-38). Hippocampus fuscus differs from H. kelloggi by having less number of tail rings (33-37), dorsal fin rays (14-17) and pectoral fin rays (14-16) (Lourie et al., 1999). In H. kelloggi, the pectoral and dorsal finrays count between 17-19 and 17-19 respectively. In H. fuscus, the tail length is much lower (52.39 ± 8.04% of total body length) than H. kelloggi, where it was 57.39 ± 8.92 (% of total body length). Hippocampus kelloggi is distinctly differs from H. fuscus in snout and head length. In H. kelloggi, the snout length is 55.01 ± 15.34 (% of head length) and head length is 19.04 ± 5.05% of the total body length but in H. fuscus, the snout length is only 53.96 ± 7.72 (% of head length) and the head length is 15.10 ± 2.83% of total body length. In H. fuscus, the coronet is somewhat more   (Lourie et al., 1999;Pollard, 1984). However, specific habitat associations have been reported for H. bargibanti (Lourie et al., 1999). Different species of Syngnathids often co-exist in the seagrass beds (Curtis and Vincent, 2005;Kendrick and Hyndes, 2003). In India, the common and widely distributed species are H. kuda, H. fuscus and H. trimaculatus. Along the Palk Bay coast, the shrimp trawl by-catch generally brings in higher quantities of H. kuda. In the Palk Bay area, four species of seahorses have been recorded. H. mohnikei may not be an Indian species and it might be come to Indian coast by drifting or by water currents (Thangaraj and Lipton, 2007). Among the three species, H. trimaculatus is to be found more in the deeper waters (>50 feet) among sponges, soft corals and sandy bottoms than in the seagrass beds. This confirms the observations of Lourie et al. (1999) and this species mostly occurring in deeper waters (>15 m). Earlier survey showed that H. kuda was the dominant species in this area (Murugan et al., 2008) and it is confirmed by the present survey. Except H. mohnikei all the five species could be collected from Gulf of Mannar whereas H. kelloggi and H. histrix were not found in Palk Bay region.
Generally seahorses are patchily distributed and occur at low densities throughout the world . In Gulf of Mannar, as like in Palk Bay, H. kuda was the dominant species. Earlier reports also suggest that H. kuda was the dominant species in the Gulf of Mannar (Lipton and Thangaraj, 2002). But, Murugan et al. (2008) reported that H. trimaculatus was the dominant species in this area. Hippocampus kelloggi and H. histrix were found exclusively in the Gulf of Mannar, and absent in the Palk smooth and uncurved than in H. kelloggi. Hippocampus ingens also has nearly the same morphological traits as H. kelloggi, but differs by having highly elevated coronet and less number (37-38) of tail rings. It was reported to occur only in the Caribbean Sea (Lourie et al., 1999). Hippocampus kelloggi has a deep head of 93.37 ± 12.94% in head length compared to other Indian seahorse species. The head depth of H. fuscus and H. kuda measures 66.05 ± 8.33 and 59.50 ± 7.50% of the head length respectively. The morphometric and meristic traits of H. kuda are distinctly differs from H. kelloggi. The total body length and trunk length of H. kuda is 119.6 ± 14.03 mm and 27.75 ± 4.22 (% of total body length) respectively, whereas in H. Kellogi,it is 191.66 ± 35.98 mm and 30.13 ± 5.65 (% of total body length) respectively. And also in H. kuda, the tail is longer (59.49 ± 6.47% of total body length) H. kelloggi (57. 39 ± 8.92% of total body length).
Hippocampus kuda can be distinguished from other seahorse species belonging the H. kuda complex as shown below. Hippocampus algiricus has double eye and cheek spines, has a restricted geographical distribution, and is found in the Eastern Atlantic off the west coast of Africa (Lourie et al., 1999). There is no possibility of its occurrence in the Indian waters. Hippocampus reidi with lesser tail rings, larger coronet and double eye spines and reported from Western Atlantic Ocean only (Lourie et al., 1999). Hippocampus fuscus is often misidentified as H. kuda. Hippocampus kuda has a larger body size and backward turned coronet. But H. fuscus's coronet is slightly raised, smooth and never bent backward. In H. kuda, the snout depth is lesser (28.91 ± 4.01% of snout length) than H. fuscus (30.22 ± 7.15% of snout length). In H. kuda, the dorsal finbase length (DF) is higher (27.13 ± 5.03% of trunk length) than H. fuscus (23.74 ± 4.36% of trunk length). In H. kuda the pectoral finbase length (PF) is also higher (34.07 ± 7.20% of head depth) than H. fuscus (33.38 ± 8.84% of head depth). The head depth of H. kuda is lower (59.50 ± 7.50% of head length) when compared with H. fuscus (66.05 ± 8.33% of head length).
The three spotted seahorse, H. trimaculatus is widely distributed throughout South India. Though they are deepwater in habitat, they brought to shore mainly by the trawlers. This species also included as one of the members of H. kuda complex. The other similar species include H. fisheri, H. fuscus and H. kuda. Since all of them possess several common characteristics, some specific features are used to distinguish each one. Hippocampus fisheri has double cheek and eye spines and two prominent nose spines, whereas in H. trimaculatus, there is a single hooklike cheek spine, sharp hook-like eye spines and no nose spines. In H. fisheri, enlarged spines are scattered throughout the body, but in H. trimaculatus, the spines are very small. Hippocampus fisheri has fewer tail rings (37-38), dorsal (17-18) and pectoral finrays (13-16) than in H. trimaculatus. Moreover, H. fisheri is found only in Hawaii waters (Lourie et al., 1999). Hippocampus trimaculatus has Bay. Of these five species, H. histrix was found in very low density in India, also occurs only in very deep sea. Lourie et al. (1999) reported that H. kelloggi and H. histrix prefer deeper waters than shallow seagrass beds.
Recently Salin et al. (2005) reported that in Kerala coast, the shrimp trawl by-catch of seahorses was dominated by H. trimaculatus. The present observation confirms Salin's report, there H. trimaculatus contributes a major share in the total seahorse by-catch. The occurrence of H. fuscus observed in Kerala coast is the first time report. The contribution of this species is extremely poor in this area; this is possibly due to the lack of suitable substratum (seagrass bed) for this shallow water species. Hippocampus histrix was not found in Malabar and also suggesting due to the geographical rigidity in their occurrence.

Conclusion
The morphometric and meristic character standardization is essential for conservation purpose; enabling effective communication among fishers, traders, scientists, fisheries officials and custom officials. Taxonomic confusion complicates assessment of species conservation status, forcing unnecessary listing of synonyms in IUCN Red List of Threatened Species and greatly increasing the possibility of omitting species definitions is growing rapidly in India. From this study it clearly denotes that H. fuscus, H. kuda and H. trimaculatus, are the commonly distributed species in all the three observed areas. The shrimp trawl by-catch was bringing abundant H. trimaculatus than the other species in Kerala coast. In the context of the current ban on seahorse exploitation and trade in India, it is important to monitor their populations regularly from all major habitats to determine their progressive revival to the optimum level.