Heroini and the like

[Willem Heijns]

I have recently come across a publication (in press) by Gustavo Concheiro and others to be published in the journal Molecular Phylogenetics and Evolution. the title is: "Phylogeny and biogeography of 91 species of heroine cichlids (Teleostei, Cichlidae) based on sequences of the Cytochrome b gene". this article will cause quite a stirr-up in the way we have looked upon these interesting cichlids so far. according to the authors the publication will be followed by taxonomic work, leading to many new (generic) names for our fish.
at this point it is important to note that the analysis is based on one mitochondrial gene. some might say that this is a very small basis on which to draw such far reaching conclusions.

in the following illustrations I have tried to summarize the cladograms from the article, adding to them my understanding of generic assignments (in yellow) and a list of species Concheiro et al did not include in their study (in blue).


here's the picture for the Heroini:

http://www.willemheijns.nl/webfoto/Concheiro2006heroini.htm

observations:
I agree with the authors to prefer the use of the generic name 'Heros' (single quotes) for species without any better generic placement over the continued use of 'Cichlasoma'. this clearly states that the species in question are part of the Heroini.

octofasciatus is nowhere near any genus used in the past i.e. Archocentrus.


and here's the amphilophines named by Concheiro

http://www.willemheijns.nl/webfoto/Concheiro2006amphilophines.htm

observations:
the name of this group is not very well chosen considering that Amphilophus labiatus (type species of Amphilophus) is not at the base of the group, nor are the species of Amphilophus dominant (at least not in number).

squeezing sajica between loisellei and dovii/managuensis doesn't appear to me as very realistic. the same can be said for centrarchus.

if myrnae is more closely related to a population of nigrofasciatus than that population is to another nigrofasciatus population that myrnae should be included in nigrofasciatus or one of the nigrofasciatus population is in need for a new name. both alternatives seem strange to me although I haven't seen any pictures (I do like to look at morphology ).

istlanus and wesseli being sister species is a challenging thought. but what about grammodes?


and the herichthyines:

http://www.willemheijns.nl/webfoto/Concheiro2006herichthyines.htm

observations:
(too) many species are missing. enigmatic species like bocourti and pearsei would have been indispensable here.

relationships within Thorichthys are unclear. what if maculipinnis is to include a population of aureus?

the type species of Theraps is not included in the study. using the generic name Theraps in the taxonomic work to follow is going to be very problematic. the placement of the species now assigned to Paraneetroplus is also of main interest.

cyanoguttatus is the type species of Herichthys. same problem as with Theraps.

any comments, questions or additions will be very much appreciated.

[michi tobler]

Hi Willem,

nice work! I think the next two years are going to be very exciting for all of us. There is a lot of work being done right now and many papers are about to be published.

Concerning the cytochrome tree: I just recently talked to some people that are using a much broader approach to look at heronie phylogeny. Cytochrome b seems to be the only marker that consistently yields in inconsistent results... but as I said there seems to be more to come soon;o)

Cheers, m

[Bas Pels]

Too bad those species were not included. Willem will remember that we had (on a Dutch forum, I'm sorry) a discussion about whether nourissati belongs in the 'longimanus group' or not. I would certainly like to know whether lentiginosum (also not included) is more related.

Regarding the yellow boxes, I see a lot of boxes which raise my eyebrows, but that does not really matter

Bas

[Don Hiatt]

"squeezing sajica between loisellei and dovii/managuensis doesn't appear to me as very realistic. the same can be said for centrarchus."

That was the first thing I noticed.

Why cyanoguttatus appeared to be excluded from the other apparently closely related species of Herichthys was another one that didn't sit well with me. That may be due to the fact that specimens were unavailable for testing.

I know I haven't seen a true cyanoguttatus for some while now, but this does stick out.

urophthalmus / splendida closely related? I noticed the sharp angles and shapes of the heads and mouths of these two species resembled each other, but I always chalked that up to being a superficial similarity.

They are found in the same biogeographic region, so I am not too surprised. But does this chart really mean anything?? I wouldn't know.

I still feel that the “C”festae group is the most likely candidate as sister group to the genus Australoheros. I have yet to see a genetic analysis that reflected a relationship between the two groups. Australoheros may be out there on their own or the methods used to determine the relationship between these species is not thorough enough yet.

Willem,
You said:
“at this point it is important to note that the analysis is based on one mitochondrial gene. some might say that this is a very small basis on which to draw such far reaching conclusions.”

I agree 100%.

Mitochondrial testing for taxonomic purposes is still fairly new. As the technology and techniques are honed, I think we will start to see more reasonable data coming from taxonomists.

[Willem Heijns]

in the meantime I've studied some more phylogenies based on Cytochrome-b (Hulsey et al 2004,2006), Martin & Bermingham 1998, Roe et al 1997). I will comment on them at a later point.

most of these phylogenies show Neetroplus nematopus and Hypsophrys nicaraguensis to be each others closest relatives. this suggests that they share a common ancestor and are thus sister species.

the conclusion that they should therefore belong to one genus is however not justified. that would be a taxonomic statement and taxonomic statements go by other rules than phylogenetics sometimes may imply.

any taxonomist working with morphologies would notice the many differences between the two species and probably never assign them to one genus.

[Willem Heijns]

I've changed the illustrations into separate webpages to make them easier to interpret. any reactions?

[Juan Artigas]

I basically feel comfortable with what Willem and Don said here. I will comment on the Heroini cladogram:

A. robertsoni derives from altifrons and then from robertsoni derives A. diquis (sympatric to A. altifrons), A. rostratus and A. longimanus.

What looks unrealistic to me is the fact that this result would probably require that A. robertsoni had an ancestral population in the ichthyological province of San Juan, that population later disappeared (or turned into A. longimanus) and the population just remained in the Usumacinta province, quite unlikely in my view. I am taking into consideration the geological history of Central America. I think something is cheesy with A. robertsoni in that graphic. The rest I feel logical in zoo geographical and morphological terms.

I always personally felt (looking at base coloration, distribution and morphology) that A. longimanus would be more closely related to the A. labiatus clade than to A. altifrons. The result of this comparison though is interesting to me.

The placing of Archocentrus octofasciatus (I am following Miller for Archocentrus) is for my understanding the strangest thing on this graphic. Splitting from A. macracanthus? Then herichthyines and festae (In a different ichthyological province) deriving from it. Herichthys we know are at least 7 million years old (When the Neovolcanic ridge closed the gap in the gulf of Mexico). So in that explanation how old would that require A. macracanthus species to be? Older than A. citrinellus? Wasn’t at that time Central America an archipelago?

Those are results that call my attention.