Comprehensive DescriptionRead full entry
[[ Genus Archocentrus Gill ]]
The traditional “working diagnosis” of Archocentrus was based strongly on meristics, especially on the high number of dorsal and anal fin elements, with formulae D. XVII-XX,8-11 and A. VII-XII,6-9. There have been also morphometric attributes, albeit rather general: the pectoral fin reaching or exceeding the anal fin origin; body somewhat deep; snout short, mouth rather small, maxilla not reaching the orbital rim, neither vertically nor horizontally; teeth in the outer row subequal in size anteriorly in both jaws (Regan 1905; Kullander 1996; Greenfield & Thomerson 1997; Miller et al. 2005). No synapomorphies were known for Archocentrus , Cryptoheros ZBK , or Hypsophrys ZBK , but some of these traditionally diagnostic characters have turned out to be synapomorphic, e.g. the number of anal-fin spines for Archocentrus ; others, as expected, were convergent, e.g. the length of the maxilla (Schmitter-Soto, in press). Still others, such as body depth, are useful for alpha-level taxonomy, but not for phylogeny reconstruction (Zelditch et al. 2005).
Allgayer (2001) had already expressed the view that Archocentrus should be restricted to Ar. centrarchus and Ar. spinosissimus . Burgess (2000) judged that Ar. multispinosus should be added too, and Allgayer (2001) himself included Herotilapia ZBK in his subtribe Archocentrina (= Archocentrus + Amatitlania + [ Cryptoheros ZBK without Cr. panamensis ], a group for which Schmitter-Soto [in press], found no support). This opinion, although only recently published explicitly, is a venerable one: Regan (1908) called Herotilapia ZBK “evidently closely allied” to Archocentrus . Other authors, e.g. Bussing (1976), have considered Herotilapia ZBK “derived” from Archocentrus .
However, most molecular phylogenetic hypotheses (Martin & Bermingham 1998; Farias et al. 2004; Hulsey et al. 2004; Concheiro Pérez et al. 2007 -all based on the cytochrome b gene) have found Herotilapia ZBK quite unrelated from Archocentrus and allies. A possible explanation for this pattern may lie on the intrinsic limitations of cytochrome b. As stated by Martin & Bermingham (1998), “ heroine cichlids… are unfortunately refractory to phylogenetic inference using cytochrome b sequences.” Long branches are common in “dwarf” Neotropical cichlids, such as species of Archocentrus and allies, and the cytochrome b gene presents saturation at third-codon position (Farias et al. 2004). Other genes should provide interesting independent evidence to settle the issue.
My understanding of Cryptoheros ZBK diverges from Allgayer’s (2001), who included also what is here called Amatitlania , and excluded Cr. panamensis . On the other hand, Kullander (2003), who did not recognize Cryptoheros ZBK , decided to call this species Archocentrus panamensis , recognizing that its affinities do not lie with Neetroplus ZBK (nor with Hypsophrys ZBK ). Am. nigrofasciata had been considered a Cryptoheros ZBK by Allgayer (2001) and an Archocentrus by Kullander (2003).
Neetroplus ZBK has had a rather unfortunate usage history. Based on a very distinctive but overrated dental character (like the tricuspid teeth of Herotilapia ZBK ), monotypic at first, it became an artificial group with the inclusion of several incisor-toothed cichlids, such as N. carpintis Jordan & Snyder, 1899 ZBK (= Herichthys carpintis ), N. bocourti Vaillant & Pellegrin, 1902 ZBK (= ‘Cichlasoma’ bocourti ), and N. panamensis ZBK (= Cr. panamensis ). Among this assemblage, H. nematopus is the only species to have incisor teeth as juvenile (Rogers 1981).
The synonymization of Neetroplus ZBK with Hypsophrys ZBK has the disadvantage of obliterating a widely used name ( Neetroplus nematopus ZBK ). However, a classification should also convey information about phylogeny, and this purpose is not served well by recognizing two monotypic sister genera instead of calling their clade by just one generic name. An example of this stance is the synonymization of Garmanella ZBK with Jordanella ZBK (Cyprinodontidae), in “…the interest of having generic categories define derived groups, rather than recognize individual differences…” (Parenti 1981). Moreover, the sister-group relationship of H. nicaraguensis and H. nematopus is supported by most proposed phylogenies (e.g. Martin & Bermingham 1998; Hulsey et al. 2004; Concheiro Pérez et al. 2007).
It should be acknowledged that a classification alternative to the one here proposed, and equally congruent with the phylogeny presented by Schmitter-Soto (in press), might include all species dealt with in this work in the genus Hypsophrys ZBK (or Amphilophus ZBK , or Parachromis , all three names having been proposed in the same work, on the same page -Kullander & Hartel 1997). It is my opinion, notwithstanding, that the classification proposed here is more informative, and also less disruptive with recent usage. On the other hand, since the aim of this work is to review the species formerly assigned to Archocentrus , the choice of taxa was not ideal for such higher level decisions.
I follow the Evolutionary Species Concept (ESC). However, the ESC is not an operational concept; to be able to recognize species under the ESC (i.e., every lineage that “maintains its identity from other such lineages and which has its own evolutionary tendencies and historical fate”: Wiley 1978), I applied here both the “autapomorphic” and the “diagnosable” versions of the Phylogenetic Species Concept (PSC) (Mayden 1997): every species should display at least one derived character of its own (Rosen 1979), but, if a non-homoplastic (unique) autapomorphy is not found, then “the smallest diagnosable cluster of individual organisms within which there is a [presumed] parental pattern of ancestry and descent” (Cracraft 1983) is to be recognized as a species. Thus, I agree with Lucena et al. (1992), who point out that “to consider two distinguishable populations as a single species [implies] loss of information about phylogeny and biogeography.”
The phylogeny and biogeography of the species formerly assigned to Archocentrus are discussed by Schmitter-Soto (in press).