Cichlidae classification

[Juan Artigas]

I think this is not a matter of distribution and species number. The separation of Ptychochrominae from Etroplinae is well supported by several studies and in my view even if one takes only a look at these fishes. Personally, I would prefer the 'four-subfamily-solution' of Sparks & Smith over a splitting into much more subfamilies for sake of clearness.

Thank you Rico, that seems to be the case.

[Juan Artigas]

Well I think we have a little more solid classification of Cichlidae now. Please take a look at at it and let me know your comments. Next step will be a page per subfamily/tribe where a type genus, diagnosis and comments are offered, plus a reference for the classification of each genus into the given tribe. Thanks for your help

The link again:

http://www.cichlidae.com/gallery/spe_classification.php

[Rico Morgenstern]

That looks quite good already. A little error: the tribe Cichlini is listed twice, the 'African one' must read Bathybatini Poll, 1986 (Trematocarini Poll, 1986 is a synonym, see Takahashi 2003, I will upload this in the CRC library). There are several more tribes established for Tanganyikan cichlids, possibly too much. The problem is that the phylogenetic studies on these fishes have been done without looking at their relationships to other African cichlids. Tropheini may be, as seems from several molecular studies (most comprehensive is Koblmüller & al. 2008), only a subgroup of Pseudocrenilabrini.

Thank you I had realized that, I had just procrastinated its removal, sames goes for Amphilophini and Astatheroini groups, which I left out for the same reason. I however believe they could be useful if properly proposed

The several published phylogenies differ in number, composition and naming of these groupings, and non has been poperly diagnosed so far. There are two apparently available names which must be taken into consideration - Therapsini Allgayer, 1989 and Archocentrina Allgayer, 2001. The former has been largely overlooked and must probably replace the tribe name Heroini Kullander, 1998 for it has priority, however, there is further research required. And if the so called Herichthyines are formally recognized as a subtribe including also the genus Theraps, the name of the subtribe would be Therapsina. Archocentrina in turn would be the oldest available name for the group called Amphilophines.

However, this would probably give too much subtribes. A more consistent subgrouping of the tribe Heroini or Therapsini, respectively, seems to be that of Heros, Symphysodon, Uaru, Pterophyllum and Mesonauta being placed into one group which would - if recognized as a subtribe - have to be called Heroina Kullander, 1998 (this is rather unfortunate for we have also a genus Heroina Kullander, 1996, however I found no regulation for such a rather unique case in the Code, at least it seems not covered by the Principle of Homonymy), another one including all North/Central American and Antillean genera as well as the South American genera Australoheros, Caquetaia, Heroina and the 'Cichlasoma' festae group (the so called Circumamazonian Heroines of Conchiero Perez & al. 2007) which would take the name Therapsina. Hoplarchus and Hypselecara may form a third group not yet named. The enigmatic 'Cichlasoma' microlepis still awaits its rediscovery and cannot be assigned yet.

Please note that this all remains speculative, especially in the absence of diagnoses. It is only my personal interpretation exposed here to discussion.

[cichla]

The contribution by you, Juan, is informative and it is an important step to obtain a robust classification. However,I think Philippe is right to mention ''... I think that you'll always find somebody to say that things are wrong because nobody exactly agrees with his neighbor''. This is why I am convinced, that the classification by Smith, Chakrabarty & Sparks will not be the final classification.

For example within the 'Evolutionary Classification' the tripe Acaroniini makes sense, but not within the 'Phylogenetic Classification'.

Subtribe yes or no? I think only the subfamily and the tribe are practicable infrafamilial levels (see Kullander 1998). Subtribes and such stuff seems more useful for 'philatelic' ones, who just like to collect names, or for PhyloCode followers, who like to name every single clade.

Therapsini: I have no doubts that Therapsini Allgayer, 1989 has priority over Heroini Kullander, 1998.

[Juan Artigas]

Thank you both Rico and Ingo, I have corrected Bathybatini. I will take a look at Therapsini and implement it instead of Heroini. As I explained before, I will just work with the sub-family and Tribe levels, leaving out other aggrupations, this is because I believe the practical use of the intrafamilar levels is better keeping it that way, and besides, we are still struggling to comprehend and general agree on our concepts of genera, tribes and subfamilies to still try to go one step further. This way we can focus our discussions and energy more efficiently.

The format in the CRC is flexible so we can always incorporate all the new research results and proposals. I will work out the tribe/sub-family pages and be back here. I have already incorporated in the database all the diagnosis for the American sub-family and tribes compiled by Smith, Chakrabarty and Sparks (2007).