[[Hypsophrys Agassiz 1859]]

Discussion

There is sufficient justification to synonymize Neetroplus and Hypsophrys because of their monotypic status and sister-group relationship, even though they do not share the features that originally delimited the taxa (e.g., spatulate teeth of Neetroplus and the single midbody spot of Hypsophrys). These features were sufficient to diagnose the included species at the time of their description, but because the original diagnostic features occur in subsequently described taxa, updated diagnoses are warranted and the synonym is justified. A sister-group relationship for these taxa was originally recovered in the molecular phylogeny of Hulsey et al. (2004) with parsimony bootstrap and Bayesian posterior probabilities of 100%. Hulsey et al. (2004) recovered a large clade of nominal Archocentrus, Parachromis and Amphilophus species as the sister-group to the Neetroplus + Hypsophrys clade (now just Hypsophrys). All species of Archocentrus, Parachromis, and Amphilophus have two supraneural bones except Archocentrus spilurus, which has one. The most closely related taxa to Archocentrus spilurus within this clade all have two supraneurals. Possession of two supraneurals is derived in most of the other outgroup Heroine taxa (Chakrabarty 2006b); therefore, the sister group to Hypsophrys would also optimize to possession of two supraneurals. Among other Heroine taxa the majority have two supraneurals; only “Cichlasoma” istlanum, “C.” salvini, and Thorichthys meeki have been reported to occasionally have one supraneural, but this feature is polymorphic (Cichocki 1976; Chakrabarty 2006b). Therefore, the diagnostic feature of a single supraneural is a synapomorphy of Hypsophrys.

It may be worthwhile in the future to investigate the taxonomic status of Cichlasoma spilotum Meek 1912, which was synonymized under Hypsophrys unimaculatus by López (1974). The synonym was presented without data from the type series or type locality. The types of Cichlasoma spilotum have a much narrower head profile than that of the holotype of Heros nicaraguensis. The paratypes of C. spilotum are the general shape of females of Hypsophrys unimaculatus. The holotype of C. spilotum (FMNH 7686; comments here are based on a photograph of the specimen) and the holotype of H. nicaraguensis are both males with nuchal humps, but the holotype of C. spilotum has a less developed hump. Nevertheless, additional material from the type locality of Cichlasoma spilotum should be obtained for study. Female specimens of Hypsophrys unimaculatus from Costa Rica presented in López (1974) generally agree in body shape to the types of C. spilotum; Miller (1966) also noted that these species may be synonyms.

Several novel behavioral features of these species are noteworthy. Coleman (1999) noted that Hypsophrys unimaculatus create horizontal tunnel nests and have non-adhesive eggs (an apparent autapomorphic feature for this species among Central American cichlids). Individuals of Hypsophrys nematopus do not build their own tunnels but have been observed occupying hollow logs and holes dug by other species (Coleman 1999). McKaye (1977) observed individuals of Hypsophrys unimaculatus guarding the fry of the piscivorous cichlid Parachromis dovii. The implication of this altruism is that survival of the predatory species would be detrimental to the altruist's chief competitor Hypsophrys nematopus. This remains a controversial interpretation (McKaye 1979; Coyne & Sohn 1978). With the taxonomic changes presented here, these putative competitive interactions should be henceforth discussed as occurring between congeners; not merely among sympatric species of the same family.