B. dorsiocellata is sometimes traded as 'green-eyed', 'emerald-eyed' or 'hi-spot' rasbora.
There appear to exist several geographical forms of the species at least two of which have been seen in the trade.
The most common is the rather plain-coloured fish that is produced commercially in large numbers and has gained popularity due to the bright greenish-blue reflective patch in the lower part of the eye.
The other is only available from time-to-time and in all likelihood is wild caught.
It's particularly attractive because the caudal peduncle and base of the caudal fin are suffused with a russet red colouration, the extent and intensity of which can vary depending on the mood, condition and sexual maturity of the fish.
B. cheeya Liao and Tan, 2011 can be told apart from B. dorsiocellata by a combination of characters as follows: lateral line complete (25-30 pored scales vs. 4-9); more scales in the lateral row (29-32 vs. 25-27); larger adult size (39 mm1.5 in SL vs. 23 mm0.9 in SL); larger head and prepectoral length (head length 28.5-30.0% SL, vs. 24.4-27.9; prepectoral length 29.6-32.9% SL, vs. 25.8-28.8). B. exilis Liao & Tan, 2014 is distinguished from the other species by the following characters: 10 circumpeduncular scales (vs. 12); anterior outline of dorsal-fin blotch ranging from 2nd unbranched ray to first four or five branched rays with transparent inter-radial membrane between 1st and 2nd unbranched rays (vs. blotch extending from 1st unbranched ray and inter-radial membrane between 1st and 2nd unbranched rays not hyaline), and relatively more slender body (body depth 19.3-24.7 % SL, vs. 24.4-26.7).
Several years prior to its description Grant (2002) suggested that specimens of B. cheeya in the aquarium trade might represent the subspecies Rasbora dorsiocellata macrophthalma (Meinken, 1951) which was described using a specimen from the aquarium trade from the vague locality 'Malay Peninsula'.
Liao and Tan (2011)[6] argue out that the type specimen of R. d. macrophthalma has been lost and that even if proved a genuine species the validity of B. cheeya would be unaffected since it possesses a complete lateral line (vs. incomplete in R. d. macrophthalma).
The genus can be distinguished by the following combination of characters: palatine process rectangular with a narrow base and lacking projection; a median black blotch in the dorsal fin; low number of predorsal vertebrae (8-9, 6:3 vs. more than 9 in other rasborin genera).
Rainboth's 'Fishes of the Cambodian Mekong' characterised members of Rasbora by possession of an unbranched, non-spiny first dorsal-fin ray and seven soft dorsal rays, origin of the dorsal fin in the middle of the body, five branched anal-fin rays, a small mouth not extending below the eye and a lack of barbels.
It has long been recognised as a polyphyletic lineage and Liao et al. (2010) proposed a number of taxonomical changes based on the results of their phylogenetic analysis.
The authors found species of rasborin genera to represent a monophyletic grouping existing in six clades, and erected four new genera (all containing former members of Rasbora) in order to preserve monophyly of the existing groups, i.e., Boraras, Horadandia, Rasbora, Rasboroides and Trigonostigma.
The first clade is monotypic, comprising only Kottelatia brittani (formerly Rasbora brittani) while the second contains only the genus Brevibora.
The third clade comprises Boraras, Horadandia, Rasboroides, Trigonostigma and three species previously included in Rasbora but moved into new genera; Trigonopoma gracile, T. pauciperforatum and Rasbosoma spilocerca.
The results for Boraras and Trigonostigma were found to be inconclusive in some respects and further work regarding their phylogenetic position was recommended.
The fourth clade includes Rasbora semilineata, R. borapetensis, R. rubrodorsalis and an undescribed fish similar to R. beauforti.
Clade five consists of the R. daniconius species assemblage, R. hubbsi, R. paucisqualis, R. wilpita, R. kobonensis, and R. ornata.
Clade six, meanwhile, is subdivided into two groupings of which the first contains R. einthovenii, R. elegans and R. cephalotaenia and the second R. lateristriata, R. argyrotaenia, R. volzii, R. paviana, R. rasbora (plus an undescribed, similar fish), R. caudimaculata and R. trilineata.
Since this latter clade contains the type species (see below) its members retain the generic name Rasbora as do clade five species because they don't differ sufficiently to warrant the erection of a separate genus.
Unfortunately many species weren't included in the analysis, meaning inevitable questions are raised regarding the correct placement of the 40 or so other Rasbora spp., in particular.
As the genus had previously been split into various 'species groups' (putative assemblages of closely-related species) dating back to Brittan (1972, who referred to them as 'species complexes') Liao et al. proposed the following arrangement whilst noting it may be subject to change with further phylogenetic studies:
R. semilineata species group: R. semilineata, R. borapetensis, R. rubrodorsalis.
R. trifasciata species group: R. trifasciata, R. amplistriga, R. bankanensis, R. dies, R. ennealepis, R. hubbsi, R. johannae, R. meinkeni, R. paucisqualis, R. rutteni, R. sarawakensis, R. taytayensis, R. tobana, R. tuberculata.
R. daniconius species group: R. daniconius, R. caverii, R. kobonensis, R. labiosa, R. ornata, R. wilpita.
R. einthovenii species group: R. einthovenii, R. cephalotaenia, R. elegans, R. jacobsoni, R. kalochroma, R. kottelati, R. nematotaenia, R. tubbi.
R. argyrotaenia species group: R. argyrotaenia, R. aprotaenia, R. aurotaenia, R. baliensis, R. borneensis, R. bunguranensis, R. dusonensis, R. evereti, R. hobelmani, R. hossi, R. lateristriata, R. laticlavia, R. leptosoma, R. philippina, R. septentrionalis, R. spilotaenia, R. steineri, R. tawarensis, R. tornieri, R. volzii.
R. sumatrana species group: R. sumatrana, R. atridorsalis, R. calliura, R. caudimaculata, R. dorsinotata, R. notura, R. paviana, R. rasbora, R. subtilis, R. trilineata, R. vulgaris.
Not classified: R. beauforti, R. chrysotaenia, R. gerlachi (validity in question), R. kalbarensis, R. reticulata, R. vulcanus (possibly not Rasboras) and R. zanzibarensis (identity in question).
Shortly afterwards a paper investigating systematics of the subfamily Danioninae, with significantly different conclusions, was published by Tang et al. 2010.
The four new genera proposed by Liao et al., plus Boraras and Trigonostigma, were synonymised with Rasbora based on an incomplete knowledge of relationships within the group, an approach described as 'more conservative', but this approach was dismissed by Kottelat (2013), who is followed here.
Editing needs JavaScript. This is the text as it was last saved.