H. kuda Bleeker 1852

English common names. Spotted Seahorse, common seahorse, estuarine seahorse, estuary seahorse, oceanic seahorse, yellow seahorse.

Synonyms. H. aterrimus Jordan and Snyder 1902, H. borboniensis Duméril 1870, H. chinensis Basilewsky 1855, H. fuscus Rüppell 1838, H. hilonis Jordan and Evermann 1903, H. horai Duncker 1926, H. melanospilos Bleeker 1854, H. moluccensis Bleeker 1852, H. novaehebudorum Fowler 1944, H. polytaenia Bleeker 1854, H. raji Whitley 1955 (= H. kuda multiannularis Raj 1941), H. rhynchomacer Duméril 1870, H. taeniops Fowler 1904, H. taeniopterus Bleeker 1852, H. tristis Castelnau 1872

Syntypes. (2) RMNH 5167 (1 of several), BMNH 1867.11.28.360 (see also Bleeker specimens: NMV 46227- 28 (2)).

Type locality. Singapore

Distribution. Australia (northern), Bahrain, Cambodia, China (including Hong Kong SAR and Province of Taiwan), Comoros, Cyprus, Djibouti, Egypt, Eritrea, Fiji, France (Réunion), India, Indonesia, Iran, Israel, Japan, Kenya, Kuwait, Lebanon, Malaysia, Madagascar, Mauritius, Mozambique, Federated States of Micronesia, New Caledonia, Oman, Pakistan, Papua New Guinea, Philippines, Qatar, Saudi Arabia, Seychelles, Singapore, Solomon Islands, Somalia, South Africa (eastern), Sri Lanka, Sudan, Syria, Tanzania, Thailand, Tonga, Turkey, United Arab Emirates, USA (Hawaii), Viet Nam.

Notes. Hippocampus kuda is a very widespread species (or species-complex) that exhibits localized haplotypes, phylogeographic structuring (Lourie 2004; Teske et al. 2005), and variable morphology. BOLD (2016) separates the 54 sequenced specimens into four BINS (Barcorde Identification Numbers) although they only differ from one another by 1.28–2.25% (648 bp, CO1), and two of the three BINS with more than a single specimen contain members of more than one purported species. Furthermore, overlapping meristics, paraphyly among purported species, genotypes from different clades (BINS) in the same populations, and lack of diagnostic morphological differences mean that, pending further research, we are unable to uphold purported species as valid in this revision. The global ‘ H. kuda -clade’ includes H. kuda, H. fuscus, H. borboniensis, H. capensis, H. algiricus, and H. reidi (Casey et al. 2004; Silveira et al. 2014; Teske et al. 2005; BOLD 2016). We have here synonymized H. fuscus and H. borboniensis due to a lack of distinguishable morphological, genetic, or geographic differences from H. kuda proper (from Southeast Asia). Note that this implies that H. kuda is in fact a Lessepsian migrant, meaning that it has passed through the Suez Canal and into the Mediterranean Sea (Golani & Fine 2002). We retain H. reidi and H. algiricus based on their subtly distinctive coronets, longer snouts, but mostly their large geographic separation (see Discussion). Further studies are needed to determine whether gene flow occurs across the Atlantic, as these two species appear to be very close genetically (1.3% divergence in 1141bp, cytb, according to Casey et al. 2004 and 1.6% divergence in 652bp, CO1, according to Silveira et al. 2014). We also conservatively retain H. capensis based on its distinctive coronet, noticeably and consistently smaller size, ecological considerations (it appears to be one of the most brackish-water tolerant seahorses and has only been found in estuaries—Lockyear et al. 2006), and conservation status (it is the only seahorse listed as Endangered on the IUCN Red List—Czembor & Bell 2012). Comparisons of cyt b sequences of present-day specimens identified as H. kuda from Hawaii with the type specimen of H. hilonis revealed the same unique haplotype and led the authors to classify them as a subspecies H. kuda hilonis (Szabó et al. 2011). That said, the Hawaiian haplotype differs from Taiwanese and Philippines haplotypes by only one and two bases, respectively. Thus we do not support the acceptance of subspecific classification. The synonymization of H. melanospilos and H. taeniopterus with H. kuda was likely Bleeker’s own (according to manuscript notes to complete Bleeker’s Atlas of Ichthyology by Popta 1895). Lourie et al. (1999) followed Popta’s synonymy and we do here as well. According to Kuiter (2009), the type specimens of H. moluccensis are housed at the Museum of Victoria, although this identification is tentative. Kuiter further identifies them as a spiny species, however after examination of a photograph of one of the specimens we conclude that it is not spiny and more strongly conforms to H. kuda (SL pers. obs.). The type description of H. moluccensis also repeatedly mentions ‘low tubercles’ and nothing about spines. The original description of H. tristis only mentions a single specimen (Castelnau 1872), however there are two type specimens in MNHN. Castelnau’s paper chiefly deals with fish from the Melbourne fish market and he gives no indication as to the origin of the specimens. The specimen labels however, suggest they are from ‘Swan River, Australia’. Both Melbourne and Swan River are outside the range of H. kuda and it is possible that the specimens came from elsewhere. Morphologically they conform to H. kuda. Other names that we synonymise, based on our examination of the type material, morphologically conform to H. kuda (e.g. H. aterrimus, H. novaehebudorum, H. polytaenia —see also notes under H. spinosissimus, H. rhynchomacer, H. taeniops) (Appendix I). Remaining names lack type specimens, or we were unable to examine the types, and are synonymised based on the original morphological descriptions (e.g. H. chinensis, H. horai, H. raji and H. taeniopterus).