Hypseleotris, Gill, 1863

Thacker, Christine & Unmack, Peter J., 2005, Phylogeny and Biogeography of the Eleotrid Genus Hypseleotris (Teleostei: Gobioidei: Eleotridae), With Redescription of H. cyprinoides, Records of the Australian Museum 57 (1), pp. 1-13 : 9

Basal Hypseleotris

The phylogenetic hypothesis presented in Figure 1 supports (with a decay index of eleven) a monophyletic Hypseleotris (clade A). The genus is diagnosed by two sexually dimorphic characters, the presence of enlarged heads in males (reversed in H. ejuncida , H. kimberleyensis , H. regalis and H. n.sp. Katherine River), and the presence of elongated dorsal and anal fins in males (reversed in H. aurea ); the presence of a distinct pigment blotch on the pectoral-fin base; a constricted pectoral-fin base; a posteriorly elongate body cavity, with 6–11 anal pterygiophores preceding the first hemal spine; a precaudal vertebral count of 13 or greater; a strongly laterally compressed head and body; a small, terminal mouth whose posterior border does not reach the anterior border of the eye; a blotch of pigment on the ventral half of the caudal peduncle (reversed in H. galii , H. sp. 3 [Murray- Darling], H. sp. 4 [Midgley’s], and H. sp. 5 [Lake’s]) and, a change in second dorsal-fin pigment to a pattern of distal stripes and proximally pale dots on black ground. The most basal clade contains the nominal taxa H. cyprinoides , H. dayi , H. tohizonae , H. leuciscus and H. guentheri . The remaining clade contains all the Australian species.

Several conclusions regarding the identity of species and the allocation of names are supported by the phylogenetic hypothesis. In addition to the well-known taxa, other names have been used for Hypseleotris species that are not Hypseleotris . Two species from China, H. compressocephalus ( Chen, 1985) and H. hainanensis Chen, 1985 , lack the diagnostic features of the genus; both lack the pectoralfin pigment blotch, the small, terminal mouth and the striped/ spotted second dorsal-fin pattern. Additionally, H. compressocephalus has dorsal fin element counts of IX+I,13, higher than any Hypseleotris except H. sp. 5 (Lake’s) ( Pan et al., 1991). Chen et al. (2002) confirm that H. hainanensis is not Hypseleotris , describing for it a new genus, Neodontobutis . The species described as H. raji Herre, 1945 is also not Hypseleotris ; it also lacks the pectoral-fin pigment blotch, the small, terminal mouth and the striped/spotted second dorsal-fin pattern, and the series of anal-fin pterygiophores preceding the first hemal spine. The locality for the H. raji holotype (CAS 139863) is listed as the Adyar River, Madras, S. India, in the records of the California Academy of Sciences. However, in the description the locality is listed as a brook near Un Long, New Territory, Hong Kong. Doug Hoese (pers. comm.) regards H. raji as a synonym of Butis koilomatodon ( Bleeker, 1849) .

Two issues surrounding the non-Australian species were considered: the number and identity of species found in the Philippines, and the validity of several taxa ( H. dayi , H. tohizonae , H. leuciscus and H. guentheri ) relative to H. cyprinoides . Herre (1927) lists five Philippine species of Hypseleotris : H. cyprinoides , H. modestus , H. agilis , H. bipartita and H. pangel . Herre (1927) doubted the presence of H. cyprinoides in the Philippines, but it has since been confirmed to exist there. Herre (1927) distinguishes H. modestus from H. cyprinoides only in sometimes having one more soft dorsal ray and one to two more anal rays. Kottelat et al. (1993) indicated that H. modestus is a synonym of H. leuciscus , itself possibly a synonym of H. cyprinoides . Examination of the type of H. leuciscus in this study (RMNH 4669) shows that H. leuciscus (and thus H. modestus ) is synonymous with H. cyprinoides , as discussed below and shown in Table 1.

The other three species considered by Herre (1927: H. agilis , H. bipartita and H. pangel ) are known only from the Philippines. Types of these species are not extant, and are presumed to have been destroyed in World War II ( Eschmeyer, 1998). Based on Herre (1927), who distinguishes them by slight differences in body proportions and by their colour patterns, all three species have meristic values consistent with H. cyprinoides . Hypseleotris agilis is the most distinctive, featuring a pattern of eight spots mediolaterally and three stripes radiating posteriad from the eye, with a fourth crossing the opercle but not extending to the eye. Specimens of Ophieleotris aporos captured at Laguna de Bay, near Manila, fit the description of H. agilis , and therefore we regard H. agilis as a synonym of O. aporos .

The two remaining Philippine species of Herre (1927) are H. bipartita and H. pangel ; the specimens Herre (1927) examined for these species fall into two size classes. Those classified as H. bipartita ranged from 22 to 37 mm in length and exhibited a spot or short bar on the ventral portion of the caudal-fin base, and a second dorsal in males with pale spots on a black background. The H. pangel specimens were 32 to 47 mm in length and featured the second dorsal, anal and caudal fins with blotches or bars, and in some specimens, a fine lateral black band. The description and illustration of H. bipartita coincides very well with that of H. cyprinoides , that of H. pangel less so, but the H. pangel specimens are larger and variations in colour pattern may develop with size. Given the similarity in colour pattern and absence of unique differences, we recommend synonymizing H. bipartita with H. cyprinoides . We reserve judgment on H. pangel ; if the colouration differences reported by Herre (1927) are not ontogenetically variable, then it may be a valid species. We have no specimens of H. pangel , therefore, we do not consider it further.

The issue of species identities throughout the range of non-Australian Hypseleotris is largely a question of the boundaries and variability within H. cyprinoides . This widespread species of Hypseleotris can tolerate salt water ( Bruton, 1996), and is known from South Africa, Madagascar, Reunion Island, east to New Guinea and north to the Philippines and Japan. The names H. dayi ( South Africa), H. tohizonae ( Madagascar) , H. leuciscus ( Indonesia and west Pacific) and H. guentheri (northern New Guinea and west Pacific) have all been used for fish that agree with the description of H. cyprinoides , with some minor variation in colour pattern. The phylogenetic analysis in this study included morphological data for H. cyprinoides , H. dayi , H. tohizonae , H. leuciscus and H. guentheri , coupled with DNA sequence for H. cyprinoides , H. dayi and H. tohizonae ; these five taxa formed a polytomy in the total evidence analysis. Based on these results we advocate synonymization of H. dayi , H. leuciscus , H. tohizonae and H. guentheri under H. cyprinoides . This recommendation is in accord with similar remarks by Hoese (1986) for H. dayi and H. tohizonae , and Kottelat et al. (1993) for H. leuciscus ( Table 1). Hypseleotris cyprinoides is formally rediagnosed below.

Figures

Fig. 1. Strict consensus of 189 most parsimonious hypotheses of Hypseleotris relationships. The numbers at nodes are decay index values, and changes of selected morphological characters are indicated with hash marks and the character number and state change (characters are represented with DELTRAN optimization).The species name and collection locality are indicated, and clades discussed in the text are lettered. In the southeastern species, abbreviations after the name and locality indicate the province or drainage basin: East Coast (EC, also known as Eastern Province), Murray-Darling Basin (MDB, Murray-Darling Province), or Lake Eyre Basin (LEB, Central Australian Province).