Synspilus A Junior Synonym of melanurus?

[Mark Smith]

Paraneetroplus synspilus is a Junior Synonym of Paraneetroplus melanurus (Teleostei: Cichlidae)
CALEB D. MCMAHAN, CHRISTOPHER M. MURRAY, AARON D. GEHEBER, CHRISTOPHER D. BOECKMAN & KYLE R. PILLER (USA)

Zootaxa 2833: 1–14 (27 Apr. 2011) Accepted: 9 Mar. 2011

[Juan Artigas]

I fully agree, although the type localities are separated just 28 km in straight line, the populations for the type specimens are probably the most distant ones in terms of distribution. So some degree of difference is expected. The diagnostic trait for P. melanurus (the angle of the longitudinal bar pointing the opercle, does not comply for several populations of P. synspillus in the eastern part of the Yucatan peninsula. Using the same criteria, I also believe that Thorichthys meeki is a junior synonym of T. affinis.

[Philippe Burnel]

Paraneetroplus ??????????????????????????

[cichla]

Dear Philippe, dear al,
yes, Paraneetroplus. See: López-Fernández, Winemiller & Honeycutt (2010): Multilocus phylogeny and rapid radiations in Neotropical cichlid fishes
(Perciformes: Cichlidae: Cichlinae). Molecular Phylogenetics and Evolution 55: 1070–1086. and McMahan, C. D., A. D. Geheber & K. R. Piller (2010): Molecular Systematics of the Enigmatic Middle American Genus Vieja (Teleostei: Cichlidae). Molecular Phylogenetics and Evolution. and the discussion here: viewtopic.php?f=17&t=9603
Greetings, IS

[Philippe Burnel]

ah yes.. cladism...

the best way seems to call every american (south and central)... Cichlasoma
and all the african cichlids ... Haplochromis

[Nuchal Man]

I was wondering if anyone could shoot me a copy of the paper. Sadly, the University I attend doesn't have access to zootaxa.

Thanks,

Sam Borstein
sam@borstein.com

[Lee Nuttall]

Interesting to note that the straight caudal bar isn't always consistant on the same specimens.

Interesting paper.

[Willem Heijns]

This was bound to happen. The variability in Cichlasoma synspilum Hubbs 1935 was big enough to encompass Heros melanurus Günther 1862. So now we have a statistical backing for this synonymy. I can change my database.
Interesting notion is that in this age of molecular analyses the matter has been decided upon using only morphological characters. Also, the authors are not fully aware of the regulations in the ICZN. Redescribing the species Heros melanurus as Paraneetroplus melanurus does not render the original name a synonym.

About generic assignment: in CLOFFSCA Kullander (2003) grouped a number of species under Vieja without any comment. Or, as McMahan et al (2010) stated: "without a re-diagnosis or formal taxonomic treatment". No wonder a genus like that turns out to be paraphyletic.
The phylogenetic tree of McMahan et al (2010) is mainly based on the mitochondrial cytochrome-b gene. Some would say that is not a very solid basis. Be that as it may, making taxonomic recomendations from is is quite arbitrary. McMahan et al propose Paraneetroplus for the majority of the former Vieja species. But if you take a closer look at the clade at hand, you can distinguish three monophyletic groups within the clade, neatly corresponding to the genera Paraneetroplus, Vieja and Paratheraps. Illustrated here with McMahan's proposal on the left side and my alternative on the right side.
No argumentation is given as to why the generic grouping was made at the higher level in the tree. It is just as subjective as my alternative which leads to the retention of at least some well known genera (my favourites of course ).

I think we're not done here yet. For now, my database will contain Vieja melanura.

[Thomas Andersen]

[quote="Willem Heijns"]But if you take a closer look at the clade at hand, you can distinguish three monophyletic groups within the clade, neatly corresponding to the genera Paraneetroplus, Vieja and Paratheraps.

Yes, but that's exactly what a strict cladist will not accept. In cladism you cannot have "monophyletic groups" within a clade, they will see it as paraphyly which is NOT ACCEPTED in cladism. In their eyes ALL descendents from a common ancestor should be included (= a whole clade) - in traditional Linnean taxonomy paraphyly is accepted, so you can have several groups/genera within a clade.

I really do not hope that cladism will be the norm in the future, to me it's a step backwards, and not forwards

[Willem Heijns]

I don't quite agree Thomas. Take a look at the left side of my drawing. All species shown form a monophyletic group. Still, McMahan et al define their genus Paraneetroplus (a monophyletic group at a lower level) next to Theraps (also a monophyletic group). My point is that they give no criteria as to why they chose this level for their generic assignments.

On a theoretical note: every species forms a monophyletic group with its sister species (if you include their common ancestor). There is monophyly on many levels.

[Rico Morgenstern]

So, synspilum is gone, and it seems that we'll have to get used to another name for a familiar species. Nevertheless, this is a sound paper - far more convincing than any previous one maintaining synspilus. And in admitting intraspecific variation instead of making new species from every marginal difference, it contradicts the spirit of our time in a really pleasant way.

I have noted two small but pardonable errors:

(1) The distribution map (fig. 1) is somewhat inaccurate in showing the total range (indicated by grey-shaded area) to include the upper Grijalva drainage, the species is (like many others) restricted in this system to lowland tributaries in the State of Tabasco.

(2) 'Paraneetroplus' maculicauda does not occur in the Usumacinta drainage as stated on page 10 (quoted from Kullander 2003). The record from the Usumacinta drainage is based on a misidentification of 'Cichlasoma' heterospilum by Rivas(1962; in the original description of Thorichthys pasionis: Quarterly Journal of the Florida Academy of Sciences 25 [2]). Miller has adopted it in his famous work "Geographical Distribution of Central American Freshwater Fishes" (Copeia 4, 1966) but corrected it in the addendum to the reprint of this paper in Thorson & al. (1976: "Investigations of the Ichthyofauna of Nicaraguan Lakes").

Also, the authors are not fully aware of the regulations in the ICZN. Redescribing the species Heros melanurus as Paraneetroplus melanurus does not render the original name a synonym.

I would admit that it is not an uncommon practice two quote the original combination in the synonymy of a redescribed species. However, the style is somewhat unusual. Normally, synonyms, misidentifications and former combinations are listed directly under the heading of the accepted name. To some extent, this is certainly attributable to Zootaxa's general shortcomings in layout and style. Be that as it may, it does not detract the value of the present paper.

As to the generic assignment, we'll have still to wait for more definite works. The results of the published phylogenetic analyses are not consistent, and proper diagnoses are not available for any concept of the genera involved. I'm not at all convinced of the distinctness of Vieja and Paratheraps. A genus Paraneetroplus with the species bulleri, gibbiceps and nebuliferus (I am well aware of the strange placement of this species in McMahan&al. 2010) would perhaps be diagnosable, but I am unable to follow the inclusion of regani and argentea in that group.

Yes, but that's exactly what a strict cladist will not accept. In cladism you cannot have "monophyletic groups" within a clade, they will see it as paraphyly which is NOT ACCEPTED in cladism. In their eyes ALL descendents from a common ancestor should be included (= a whole clade) - in traditional Linnean taxonomy paraphyly is accepted, so you can have several groups/genera within a clade.

Interesting, but from this point of view one could assign all the species in the tree to Herichthys. I agree with Willem, it's only a matter of where you draw the 'line' of the genus category The genera of Willem's own variant would be monophyletic, as would be Paraneetroplus in McMahan's variant. However, the unification of the Vieja and Paratheraps clades to one genus would make that genus paraphyletic with regard to Paraneetroplus.

But why couldn't paraphyletic groups be natural. The dogma, that genera (and species) have to be monophyletic, will eventually lead either to excessive splitting or to lumping into uninformative taxa. An (simplified) example: The species flock of Lake Barombi Mbo in Cameroon consists of eleven species which are so distinct that they have been assigned to five genera, four of them endemic. The species which gave rise to this assemblage was identified as the widespread Sarotherodon galilaeus galilaeus (Schliewen 2005). So this species is paraphyletic with regard the Barombi Mbo flock. The removal of the paraphyly would require either to synonymize all the Barombi species with S. galilaeus galilaeus or to splitthe latter into an indefinite number of indistinguishable species and genera.

[Thomas Andersen]

Sorry Willem, I was assuming that the expanded Paraneetroplus, Herichthys and Theraps each represented a clade.

But why couldn't paraphyletic groups be natural. The dogma, that genera (and species) have to be monophyletic, will eventually lead either to excessive splitting or to lumping into uninformative taxa.

Exactly, that is what I fear! If all taxa should be grouped into expanded genera representing whole clades, it becomes uninformative, as it says nothing on the relationship at the species level - of course some species are closer related to each other than to others within a clade, which should be grouped into genera, but the cladistics do not share this view. I thought this was what happend to Vieja and Paratheraps

[Bas Pels]

But why couldn't paraphyletic groups be natural. The dogma, that genera (and species) have to be monophyletic, will eventually lead either to excessive splitting or to lumping into uninformative taxa. An (simplified) example: The species flock of Lake Barombi Mbo in Cameroon consists of eleven species which are so distinct that they have been assigned to five genera, four of them endemic. The species which gave rise to this assemblage was identified as the widespread Sarotherodon galilaeus galilaeus (Schliewen 2005). So this species is paraphyletic with regard the Barombi Mbo flock. The removal of the paraphyly would require either to synonymize all the Barombi species with S. galilaeus galilaeus or to splitthe latter into an indefinite number of indistinguishable species and genera.

Good example. reading the above I was thinking along the same liines, but could not produce one

So in this discussion, we tend to conclude a paraphyletic group should be acceptable - assuming it is a parent group to the group(s) which descendant from it and made the group parafyletic

[Willem Heijns]

A real systematist's dilemma! And not very well understood by everyone. I'll try and work out an example here.



Above is the classical three taxon statement used by cladists. It means that the species in group 2 and 3 are more closely related to each other than any of them is with the species in group 1. Straightforward isn't it? Now suppose we want to give generic names to our three groups. Taking into account that genera should (ideally) be natural (i.e. monophyletic) groups we could arrange them into two genera A and B as shown here:



That would make sense. By the way: paraphyletic groups are by definition not natural. The real question is: should named genera always be natural groups?

The three species groups shown here are most likely not new to science. In fact I took this example from Musilová et al (2009). The current generic assignment of our three groups is as follows:



Now here's the problem. Our genus Aequidens suddenly has become paraphyletic (not all descendents are included). How can we solve this problem that only arose when we found out about the true phylogenetic interrelationship of these groups?
There are two possiblities. One (proposed by Musilová et al) is to group them all together in one genus. That genus would be called Cichlasoma, because that is the oldest available name for the species in the group. Aequidens is then declared synonym to Cichlasoma.
The other possibilty is to leave Cichlasoma as it is and create a new genus for either group 1 or group 2 (depending on where the type species of Aequidens is in).
Either way, generic names will change but all genera will remain monophyletic.

Of course we could change nothing. Then Aequidens would become a paraphyletic genus. Is that a problem? Maybe not, but whenever phylogenies have to be presented we would need an informal name for one (or both) Aequidens groups. Phylogenetic relationships between genera would be difficult to show.

So the dilemma is: what price are we prepared to pay for stabilty in nomenclature? Or, to put it another way: how can a system aimed at stability reflect the ever growing knowledge of our cichlids?

[Bas Pels]

By the way: paraphyletic groups are by definition not natural.

To me, this is a statement which needs explanation

As Rico explained, Sarotherodon galilaeus galilaeus (Schliewen 2005) gave rise to the species flock in Lake Barombi Mbo in Cameroon.

In this time, the othere S galilaeus did not change, and therefore the group S galilaeus is still the same group.

Just because some descendents changed does not mean the other populations became new species.

By now I have another example - the species flocks in the craterlakes in Nicaragua. As far as I know - I'm not certain - they descended from Amphilophus citrinellum - but nothing changed for A citrinellum, did it? Only those which became species flocks changed

[Willem Heijns]

A natural group is the same thing as a monophyletic group and is defined as "a group containing a common ancestor and all of its descendents". Any other group (i.e. paraphyletic or polyphyletic) cannot be a natural group.

As for ancestors: when a population of a certain species is (reproductively) split into two and each of the resulting populations follows its own evolutionary path, two new species might arise. Cladists would say that the ancestor becomes extinct as a result of this. There is logic in there. Take Laguna Xiloá. At some point in time a "citrinellustype" cichlid colonized this crater lake. In the isolated environment three new species arose (amarillo, sagittae and xiloaensis). They may look a lot like what we call citrinellus (from other locations) now, but that does not make citrinellus their ancestor. Say the event took place some 10,000 years ago. Was the population in Lago Managua (which we now call citrinellus) the same then? Or has it changed just as much as the fish isolated in Laguna Xiloá?

Here's another systematist's dilemma: "can ancestors survive speciation events?" Cladists clearly say "no" to this question. And more practically: how much change is needed for each of the two populations (just been split from one) to be called a new species? Which of the two gets the new name?

[Philippe Burnel]

Here's another systematist's dilemma: "can ancestors survive speciation events?" Cladists clearly say "no" to this question. ?

And why ? I don't see any reason to beleive that ancestors don't survive. And I'm pretty sure that they do (not always but sometimes)

[Tachymarptis]

Anyway, no matter if the species (or any taxon of higher level) survives or not. Our linnean nomenclature is supposed to classify all organisms, recent or extinct. Species, genera, families, and so on follow each other for billions of years. To classify them only as clades is simply a nonsense.

Even if we do not take into account extinct taxa, it is untrue to suppose that living ones change at the same rate (the Barombi-Mbo species flock is a good example). This is one of the oversimplifications of cladism which do not correspond to the reality. Another of these oversimplifications is that the "living tree" is not necessarily a tree, but very often a reticulum, and nodes are not points in a mathematical sense (without any dimension). These schematisations are only approximations, not the reality.

So, cladistics must only be considered as useful tool, not a dogma; and parapyhyletic taxa are unavoidable in a linnean classification.

[Willem Heijns]

@ Philippe: extinction of ancestors is part of the theoretical concept of cladism. Assuming a species is defined by its characters and a certain amount of variabilty a speciation event changes that for both resulting species. Hence the "old" species ceases to exist.

@ Patrick: you are clearly not a cladist. I find much in cladism that helps me understand biological history, but I agree we shouldn't treat it as a dogma. Neither should we declare the Linnean system as the best (only?) way of classifying organisms. And approximations are the best we can get when it comes to deciphering history, aren't they?

[Tachymarptis]

Willem, I was filled with enthusiasm when cladism began to be widely teached, in the 80's. But I changed my mind when I realized step by step that their sycophants took their reasoning to reach the absurdities which I mention. Some of them claim unhonestly that classifications were not phylogenetical prior to them. This is not true: avoiding polyphyletic groupings was the aim of taxonomists since the 19th century (though not always successfully, of course). Claiming that only clades are monophyletic groupings is already unhonest: monophyletic means from a single ancestor, fullstop. A well-defined paraphyletic group is from a single ancestor, no matter if it includes all its descendants or not. In linnean classification, clades are only possible for taxa which are located at the end of a branch. Otherwise, for example, not only we should reassign most neotropical species to one genus (Cichlasoma), but also to one single species. That's the main drawback of this "only clades" dogma: you are constantly forced to "blackhole" all taxa inside the older one to respect this dogma.

I agree with you that cladism helps understanding the evolution of the living, that's why I wrote that it was a good tool; not a good master. Also, linnean nomenclature is a good tool for our human mind to understand the living and to fix its history in mind. You may find a mathematical model (for example, based on numbers or functions) which will be more rigorous for defining the evolution of the living organisms, but it will be less eloquent for our understanding. So, in my opinion, cladistics must help to define linnean taxa (with the constraints linked to them) as faithfully as possible, but must not be the aim by themselves.