Archocentrus vs. Cryptoheros

The classification of Regan (1905)

The classification used by Kullander (2003) was originally proposed by Regan (1905) as his Section 5. More information of the current classification can be found in the "Short synopsis of Archocentrus".

Alternative classification of Allgayer (2001)

Allgayer (2001) pointed out that Archocentrus sensu Regan is polyphyletic and proposed an alternative classification by describing the genus Cryptoheros (type species: H. spilurus):

Archocentrus A. centrarchus A. spinosissimus

Cryptoheros The black group C. nigrofasciatus C. sajica C. spilurus The yellow group C. altoflavus C. myrnae C. nanoluteus C. septemfasciatus

According to Allgayer (2001), Cryptoheros differs from Archocentrus in having a more slender body; slightly reduced spines in the anal fin (6-10 in C. vs. 9-12 in A.); a distinct color pattern in the dorsalis of females; and the habit of breeding in crevices.

The validity of Cryptoheros sensu Allgayer was questioned several times (e.g. Kullander; Tobler, 2004). The main issues with Allgayers's classification presented so far include the limited taxon sampling and the unresolved phylogenetic relationships as well as the lack of clear autapomorphies for Archocentrus sensu stricto or lato, or Cryptoheros (although this proved to be a difficult task for many Central American genera). Furthermore, there are several studies using molecular approaches to determine the evolutionary history of Central American cichlids (e.g. Martin & Bermingham, 1998; Hulsey et al., 2003; Concheiro Perez, personal communication), all of which do not support the classification of Allgayer. The same is true for data gained by classical morphological data (Kullander et al., personal communication). Recent observations revealed differences in reproductive behavior between Cryptoheros species. It is unclear so far, if and how this finding affects the proposed classifications.

A molecular approach?

Concheiro Perez (and personal communications) presents a whole different view of the current classification. He based his view on molecular phylogenies (of the cytochrome b gene). These data indicate that Cryptoheros might indeed be a valid genus, however, restricted to the type species H. spilurus. Furthermore, Concheiro Perez claims that A. centrarchus may in fact be a small Amphilophus (s.s.) and thus Archocentrus is a synonym to Amphilophus. The other species of Archocentrus (sensu Regan) appear scattered over a wide spectrum of other Central American cichlids and many of the proposed relationships do not make a lot of sense from a biogeographical and morphological point of view. These ideas are far from conclusive (the actual data is not even formally published). Nonetheless they indicate that our current understanding of the evolutionary history (of this particular species group) is still fragmentary and may not reflect what actually happened. Furthermore, these results provide new hypotheses that should be tested and falsified rather than rejected from the beginning.