The family Cyprinidae, the largest freshwater fish family, comprises about 210 genera and 2010 species (Nelson, 1994). Our understanding of cyprinid diversity is reasonably well known, but interrelationships between genera, tribes, and subfamilies are poorly known. Many papers on cyprinid systematics at generic, tribal, and subfamily levels have been written (e.g., Banarescu & Nalbant, 1973; Cavender & Coburn, 1992; Chen et al., 1984; Gosline, 1978; Henzel, 1970; Hosoya, 1986; Howes, 1991; Mayden, 1989, 1991; Nelson, 1994; Rainboth, 1991; Wu et al., 1964, 1977). However, in most, a very limited part of the taxa was examined. This has left relationships of these taxa in a chaotic state. One of the reasons for systematic confusion at the higher taxon levels has been that the monophyly of these taxa has not been well studied, leading to different classifications. Recently, the molecular phylogenetics of Eurasian, North American, and African cyprinids have been reported (Broughton & Gold, 2000; Durand et al., 2002; Simons & Mayden, 1998; Tsigenopoulos et al., 2002; Xiao et al., 2001; Zardoya & Doadrio, 1999), however, their scope was limited as they did not include enough representative species, making relationships between higher taxa difficult to determin.
As a small step to elucidating cyprinid interrelationships at generic, tribal, and subfamily levels, we have studied bitterlings, the subfamily Acheilognathinae, which comprise three genera and more than 45 species/subspecies. They are characterized by small size, insertion of eggs into the branchial lamella of freshwater unioid mussels, and a small number of diploid chromosomes (Arai, 1978, 1988; Arai & Akai, 1988; Arai et al., 2001; Okazaki et al., 2001; Ueda et al., 2001). With the exception of the European bitterling, R. sericeus amarus (Bloch), bitterlings are naturally distributed in Asia. Of all cyprinid subfamilies, Acheilognathinae has the largest number of synapomorphies (Cavender & Coburn, 1992). This fact and the molecular phylogeny of bitterlings based on 12S rDNA (Okazaki et al., 2001) suggest their monophyly.
Morphology of the fish lateral line system has provided a rich basis for studies in comparative anatomy and systematics (Coombs et al., 1988; Hensel, 1978; Lekander, 1949; Nelson, 1972; Northcutt, 1989; Webb, 1989a, 1989b). In these last two papers, Webb has shown that patterns of lateral-line morphogenesis and its systematic relationships suggest that heterochrony is a major mechanism of morphological diversification, and has proposed that the incomplete trunk canal may be derived from the complete trunk canal by paedomorphic reduction of the posterior section of the canal.
Cyprinid cephalic sensory canals have been reported in many papers (see Table 2). As for the subfamily Acheilognathinae, Cavender & Coburn (1992), Chen et al. (1984), and Gosline (1974) have studied, but only a small number of species were examined, and only for interrelationships among cyprinid subfamilies.
Accordingly, in order to clarify relationships between patterns of the lateral line system and heterochrony in the family Cyprinidae, both the lateral line system and infraorbital bones of all 3 genera and 27 species/subspecies of the Acheilognathinae were examined.