To be or not to (apart)?

New cichlid species and taxonomy
User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Fri Jul 26, 2013 5:22 pm

Why avoid the term vicariance? After all, your biogeographic event (leading to a split) is just that.

I was quite convinced by the sympatric speciation in Laguna Xiloá. Weren't you?
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 468
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: To be or not to (apart)?

Post by cichla » Sat Jul 27, 2013 7:08 am

Yes, you are right, Willem. Vicariance is the proper term for it. However, does it cover 'speciation by niche conservatism' too'?
Willem Heijns wrote: I was quite convinced by the sympatric speciation in Laguna Xiloá. Weren't you?

Yes, more and more ('presumably') evidences for sympatric speciation published in dozen of papers every year. However, I am not sure .... perhaps the 'sympatric speciation in Laguna Xiloá' is a kind of parapatric speciation. For some this 'species' are 'just' eco-types. So, we are finally back to the question 'what is a species'? ;-)

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Sun Jul 28, 2013 8:47 am

Speciation by niche conservatism. That's a new one for me. And I guess for many people, even experts. I just read a short introduction by someone who changes the term from conservatism to conservation, while discussing it. :? I must admit, I don't see any possibility for speciation because of it. But maybe that is worth another topic?

Parapatric speciation occurs when a barrier is not complete and a (narrow) hybrid zone is present between the two separated populations. I have spent many hours diving in Laguna Xiloá, but I have seen no sign of any barrier whatsoever. It would be hardly possible in such a small lake anyway. So the origin of amarillo, sagittae and xiloaensis can only be explained (as has been done) by habitat selection, food preference and assortative mating. Which of course leaves the question if speciation is complete here, or if only variants (subspecies?) have developed. I must have some video somewhere of a mixed pair of sagittae and xiloaensis...... 8)
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 468
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: To be or not to (apart)?

Post by cichla » Sun Jul 28, 2013 1:26 pm

I think, there is a fuzzy line between parapatric and sympatric speciation. It seems obviously – at least to me – that there is no consensus about the definition of such terms. So, I think it would push the boundaries of this topic to far if I would explain here what I call 'parapatric speciation'.
Willem Heijns wrote:So the origin of amarillo, sagittae and xiloaensis can only be explained (as has been done) by habitat selection
Yes, this is what I would call speciation by parapatry. ;-)

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Sun Jul 28, 2013 2:02 pm

I don't think it is as "fuzzy" as you say it is Ingo. At least not in the case of Laguna Xiloá. In parapatric speciation the barrier is an external factor (albeit not "watertight"). The fish have no choice. Habitat selection is an "active" choice. In Xiloá, the choice is between algae covered areas, rocky habitats and sandy areas. All these are right next to each other and can be (and in fact are) easily crossed. For some reason, based within the cichlids, some have chosen the rocky habitats AND snails as food AND mating with "likeminded" cichlids. And they evolved to become xiloaensis. Others have chosen the sandy areas AND small fish as food AND assortative mating and thus became sagittae. Amarillo is no different story.

So to me, the difference between parapatric and sympatric speciation really is the active role of the cichlids themselves (in sympatric speciation) versus the dependence on external barriers upon which the cichlids have no influence at all.
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 468
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: To be or not to (apart)?

Post by cichla » Sun Jul 28, 2013 2:30 pm

I see, thank yoo Willem. Cichlid fish seems to be very smart.

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Mon Aug 05, 2013 8:16 am

At this point it is maybe a good idea to return to the original statement. Kunz' species concept holds that whenever geneflow in a population is disrupted new species will have orginated, whether they can be diagnosed or not. The weak point in this concept is that we can never be sure if this disruption is permanent or not. And of course it is very difficult to determine if there (still) is geneflow. On our human timescale there is no such statement as "no geneflow this morning, maybe tomorrow". And on the timescale that counts (much longer than our lifespans?) there is always the possibility of restoration of the geneflow. I sure like to know what Kunz would have to say about a situation where geneflow has been disrupted, no differentiation has occurred and then geneflow is restored. How many species would we have after such a course of events?

In my view, allopatry is only the start of the (potential) differentiation. But that raises the question: how much time/differentiation is needed to split the old species into (one or) two new species? This is where diagnosability comes into play. Kullander (2000) following Cracraft (1983, 1989), has promoted the phylogenetic species concept, which holds "a phylogenetic species is an irreducable (basal) cluster of organisms, diagnosably distinct from other such clusters, and within which there is a parental pattern of ancestry and descent". Strange enough, the diagnosability has gotten much more emphasis than the parental pattern, while both are needed to define a (phylogenetic) species. Be that as it may, what about diagnosability? Most of the diagnoses I have seen are purported to distinguish a species from other species, usually closely related ones. Characters are chosen accordingly. Species A diagnosed against species B would probably be done with another set of characters than species A diagnosed against species C. None of these two character sets would by itself be sufficient to establish species A. They are descriptive and that is what they should be.

Funny enough, diagnoses are required by the ICZN. We must keep in mind though that the Code is about nomenclature and not about taxonomy. All the Code does is regulate the naming of otherwise established taxa. There simply (sadly?) are no rules for establishing taxa.
On a final note, the Code was written in a time where species concepts were hardly thought about. The whole type concept (on the species level) is a clear illustration of this. A remnant of the "typological species concept"?
Slàinte mhath!

Uilleam

Rico Morgenstern
CichlidRoom Expert
Posts: 387
Joined: Sun Jul 22, 2007 4:01 am
Location: Germany

Re: To be or not to (apart)?

Post by Rico Morgenstern » Sat Aug 17, 2013 2:57 am

Willem Heijns wrote:On a final note, the Code was written in a time where species concepts were hardly thought about. The whole type concept (on the species level) is a clear illustration of this. A remnant of the "typological species concept"?
I don't think so Willem, especially since the importance of types has increased in the course of development of the nemenclature rules. As you have repeatedly and correctly pointed out, the Code is about nomenclature, not taxonomy. Accordingly, the function of types is essentially a nomenclatural one.
1.1. Statement of the Principle of Typification. Each nominal taxon in the family, genus or species groups has actually or potentially a name-bearing type. The fixation of the name-bearing type of a nominal taxon provides the objective standard of reference for the application of the name it bears.

61.1.1. No matter how the boundaries of a taxonomic taxon may vary in the opinion of zoologists the valid name of such a taxon is determined [Art. 23.3] from the name-bearing type(s) considered to belong within those boundaries.
A type is thus nothing more or less than a tool to determine objectively (in combination with other provisions of the Code), which name the taxon takes. It does not (and cannot) define the taxon. A type specimen does not necessarily represent the species it belongs to properly. This was even known to authors who did apply the 'typological species concept' (whitout knowing, for this is a 'modern' term to explain how they recognized species) in times when the principles of typification where not even formulated. A good example is the following statement from Günther (1868: 486; my boldface) in his redescription of Heros irregularis:
I have now before me numerous examples of this species from the Rivers Chisoy, San Geronimo, and Santa Isabel; and finding that the anal spines are normally five in number, the number four of the typical specimen being merely accidental, I do not
hesitate to reunite the genus Theraps with Heros.

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Sat Aug 17, 2013 4:18 am

You're quite right Rico. What I meant to say was that the typification principle in the Code may be a leftover (remnant) of the way species were looked upon in those days (and before). If I'm correct, the typological species concept (although of course not known then under that name) compares organisms to a certain type or "ideal representative" or maybe even the "essence" of the species the organism was supposed to belong to. Your example of Günther illustrates this very well.

In modern use, we know that species are variable. The type(s) should fit within this variation space. And to determine if an organism belongs to the same species, all we have to do is compare it to the type(s) and decide whether it is conspecific with it. If so, we can use the same name for it.
But this doesn't sound logical to me at all. If our type is at one end of the variation space and the organism at hand is at the other end, they may still be conspecific. They may differ substantially from each other, but still belong to the same species. This decision can only be made knowing the variation space of the species at hand. This information can/should be found in the description. In fact, the decision can be even made without using the type(s) and relying on the description only. So I'm not so sure the types are that important in nomenclature.
Slàinte mhath!

Uilleam

Ganapoes
Posts: 8
Joined: Wed Nov 16, 2011 1:54 am
Contact:

Re: To be or not to (apart)?

Post by Ganapoes » Tue Aug 20, 2013 2:21 am

would this lead to new species? And why?

User avatar
cichla
CichlidRoom Expert
Posts: 468
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: To be or not to (apart)?

Post by cichla » Wed Aug 21, 2013 7:23 am

Willem Heijns wrote: ''typification principle in the Code may be a leftover...of the way species were looked upon in those days ... ''
Well, so far I understand the historical sources: there were indeed two different opinions just before the ''precursor'' of the current Code of Zoological Nomenclature appears. The more 'typological' and the modern pre-evolutionary thoughts. Even in this very first stage, the progressive faction (mind it was even before Darwin published his most influential book) 'win' (if I understand the historical and the secondary sources correct). So, I am not sure whether the 'type specimens' and 'diagnosis' rules are really effected by this 'typological' thoughts or not. Perhaps, these rules are just there to have a prove (type-specimen) and to force authors to improve statements how the presumed new taxon differs from those which are already named.
Ganapoes wrote:would this lead to new species? And why?
if we treat every allopatric population as a different species it will lead ''us'' to name many 'new' species, for sure. This, however, happened already. It is called 'taxonomic inflation'. :shock:

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Fri Aug 23, 2013 3:03 am

I'm not sure either, Ingo. But it is interesting to read your statement about "proof" and "forced improved diagnoses". Do types really function as proof of taxa? And are better diagnoses not on the taxonomic level only? There are many "bad" diagnoses still accepted by the Code. And even now a diagnosis only has to "purport" to differentiate the taxon. Not many authors will be forced to improve their descriptions I fear. Not by the Code anyway.
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 468
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: To be or not to (apart)?

Post by cichla » Fri Aug 23, 2013 8:21 am

Thabnk you, Willem.
I re-read (skim through) some papers about the subject. It seems that the sources are drawing an ambivalent view. Anyway....

"proof": the dictionary is given more than 10 meanings of the word. I meant it like.... ''Evidence ... sufficient to induce belief“ / „Anything that serves to convince the mind of the truth“... In other words, the type(s) are there to show that there is something. Otherwise everyone could just say ... 'I saw a new species of Flag Cichlid, I call it Mesonauta xyz'. Or take such fantasy creatures like 'Nessie'. The type in the sense of the code (rules) is not to be treated as a model or a typical specimen of a taxon. It is (just) there to have a 'proof' (an evidence).

Surely, there are still species taxa which are described and treated as valid under the Code where no type specimen is deposited. In one case – believe it or not – even a woman's handbag was used as a type specimen.

Diagnoses: I agree that there are many "bad" diagnoses. I call diagnosis as 'bad' if I have to read things like 'the new species is bigger than the others'; 'the new species looks different' &c. Such diagnoses are still published. Not by the leading journals of course. Such species taxa are actually unscientific, but even those taxa are treated as valid under the code.

That such 'unscientific' diagnoses exist doesn't mean, however, that the founders of the code made a mistake. It just mean that authors who published such uninformative stuff are 'unscientific'.

Rico Morgenstern
CichlidRoom Expert
Posts: 387
Joined: Sun Jul 22, 2007 4:01 am
Location: Germany

Re: To be or not to (apart)?

Post by Rico Morgenstern » Fri Aug 23, 2013 12:07 pm

Although it might be somewhat off-topic, I would add some more remarks about the purpose of types.

I don't think that they are meant as proof for a named species does actually exist. Likewise, they do certainly not incarnate the 'ideal representative' of a species. If you have enough specimens of a species to be named, you will certainly select an average or otherwise easily identifiable specimen to serve as holotype, but the criteria remain always subjective. Subsequent findings may still render your beautiful type specimen 'untypical' for the species (apart from the fact that a single specimen can hardly be representative for a species at all).

As noted above, types have a nomenclatural function. The purpose of a name-bearing type is to link a name unambigously to a taxon, i.e. that taxon, to which the type specimen is found to belong. In practise, this purpose is not always fullfilled, for the type may be unidentifiable or, in the case of syntypes, these may belong to different species. However, the Code offers several provisions to solve these problems (e.g. by selection of a lectotype or, under certain, very strict criteria a neotype). If this is not possible or desirable, you will have simply an inapplicable name (a nomen dubium) - no need to bother with it.
Willem Heijns wrote:In modern use, we know that species are variable. The type(s) should fit within this variation space. And to determine if an organism belongs to the same species, all we have to do is compare it to the type(s) and decide whether it is conspecific with it. If so, we can use the same name for it.
But this doesn't sound logical to me at all. If our type is at one end of the variation space and the organism at hand is at the other end, they may still be conspecific. They may differ substantially from each other, but still belong to the same species. This decision can only be made knowing the variation space of the species at hand. This information can/should be found in the description. In fact, the decision can be even made without using the type(s) and relying on the description only. So I'm not so sure the types are that important in nomenclature.
But what happens if a variable species described as such is later found to be in fact a composite of several similar species? How would we determine, which of them would take the original name, when there was no type? That is, what name-bearing types actually are for. They ensure that scientific names remain applicable, regardless how the boundaries of a taxon change. The name-bearing function is not at all affected by intraspecific variability. The limits of species are determined by taxonomists (whether they apply a certain species concept or follow their intuition or whatever), and if a name-bearing type is found to fall within these limits, the name applies to that species. Again, the type does not define the species.

Therefore, you are absolutely right when you say that a type comparision is not necessarily the best way to identify an organism. Once the taxonomic identity of a type specimen is sufficiently well established (which, in itself, is also often possible without actual examination of the type), it may peacefully rest in its jar until new questions arise, which require re-examination. In this respect, types are often overrated or even misused. A mere statement like "Examination of the type material revealed, that...." seems sometimes to count more than actual consideration of characters, distributional data, etc.. On the other hand, some authors tend to denigrate taxonomic works/statements of others because no type material was examined, regardless if it was necessary for the topic at hand or not. Without claiming to judge, I must say that, if the latter is the case, it tells more about the detractor than about the criticized author.

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Sat Aug 24, 2013 5:57 am

I can understand why we have a certain need to have a physical object (a "type") with our species descriptions. But we can just as well do without them. Here's a picture taken (and slightly amended) from my Cichlasoma presentation. It shows some of the type concept and its use in nomenclature.

Image

To the right of the picture is the word diagnosis. A diagnosis is applied to species level taxa and describes the variety, clustered under one specific name. The type (one or several specimens) should fit within that variety. In earlier days species were described with a vague or incomplete diagnosis or sometimes even without one. For those species is was convenient to have a type. Identifying other cichlids could be accomplished, for lack of a proper description, using the type. Provided of course it is in good condition. Sometimes types can be unidentifiable. Nowadays we have extensive diagnoses, describing the variety included in the species. Characters used in a diagnosis can be either morphological, describe live coloration and/or concern behaviour. Identifying a cichlid can be done using the diagnosis only. As a matter of fact characters describing coloration or behaviour can hardly (colours) or not at all (behaviour) be observed in a type.

Rico states that the function of a type is to safeguard the name of a species about to be split into several new ones. But why? The original species diagnosis will be substantially changed and new ones will be published. We could agree to keep the original name for one of the species (or maybe even discard it). After all, the species originally established by its "broad" diagnosis seizes to exist and is replaced by species with much "narrower" diagnoses. Why keep the name of a once highly variable species and use it for a species with much less variation?
Slàinte mhath!

Uilleam

Bas Pels
Posts: 2260
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: To be or not to (apart)?

Post by Bas Pels » Sat Aug 24, 2013 7:53 am

Willem Heijns wrote:Why keep the name of a once highly variable species and use it for a species with much less variation?
This question implies a certain answer to the question 'what is a species'

If one admits a species is man made, the question is a valid one. However, one can also define a species as 'the conspecifics of the type material' and admit the previous vies was plainly wrong. In this case, Willems question is without meaning.

The second point of view has been, I think, the prevalent one, and therefore the question had never been asked

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Sat Aug 24, 2013 10:21 am

The issue whether a species is manmade or really exists in nature is completely irrelevant for the question at hand. My question only deals with how we recognise species. Lumpers will never think of splitting our "clustered variation" into more than one species; splitters are more eager to do so.

And saying that species can be defined as the conspecifics of the type material is turning scientific methods upside down. I'm sure this is not the prevalent view. Types only support the species and sometimes very badly so.

But I'm very proud to have asked a question nobody has ever asked before. :-P
Slàinte mhath!

Uilleam

Rico Morgenstern
CichlidRoom Expert
Posts: 387
Joined: Sun Jul 22, 2007 4:01 am
Location: Germany

Re: To be or not to (apart)?

Post by Rico Morgenstern » Sat Aug 24, 2013 11:30 am

Willem, I agree that species cannot be defined as being 'conspecific with the type material', although it is apparently sometimes handled exactly this way.
However, it is not entirely clear to me what you mean by saying that types support species?
Willem Heijns wrote:Rico states that the function of a type is to safeguard the name of a species about to be split into several new ones. But why?
Because there is no real alternative so far, unless, of course, we repeal zoological nomenclature entirely. Linking a name only with a diagnosis is futile, because diagnoses remain always subjective. What would you do if a species is to split up, starting a poll which portion would take the original name? Or if the original name is discarded, and a later author found that the split up was unwarranted, what then? Another new name? Reinstatement of the orignal name? No, the principle of typification works well and is irreplacible (at least for the moment), if we don't want to end up in 'anarchy' and chaos.

Please bear in mind, a name-bearing type in the species group has no other function than that in the genus group (type species) or in the family group (type genus). But still you insist in giving a taxonomical meaning to it. Once again, a type is not intended as an 'ultimate standard' for the species or as reference object for identification. Some type specimen may well be useful for that purpose, but it is not what they are for. Defining and recognizing species is not the same than naming them, even if the proposal of a new name requires a definition or diagnosis. A species is in fact named by identifying a type specimen as falling within its defined limits, not vice versa.

One problem is of course, that a type specimen may not show the characters required for proper identification. It may be a malformed or teratological specimen. Meristic/morphometric characters may fall into the overlapping range between two closely related species, or it may be a juvenile with the crucial characters not yet being developed. In addition, type specimens may be partially decayed or damaged; and of course a decolored specimen is useless for an identification by means of color pattern. Finally, the type material may be lost or destroyed, or indeed, it may never have been preserved. Nevertheless, it is often possible to identify it: Original (and other earlier) descriptions and figures may be useful to recognize specimens or characters subsequently lost. Of enormous help can be the type locality. Even if imprecisely or even wrongly stated in the original description, data about the collector and/or historical circumstances may enable one to locate or at least narrow down the probable origin of the type(s). All this requires some detective work but can be quite fascinating. In those cases, where the type remains unidentifiable, the name is inapplicable and we don't need to care for it until further information is available.

As a concluding note, I would say the type concept poses no inconveniences or problems serious enough to call for its abolishment. Just take type specimens for what they are, and all will be fine :wink:

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 799
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: To be or not to (apart)?

Post by Willem Heijns » Sun Aug 25, 2013 2:48 am

You know I don't give up easily Rico. :D

You give examples of how a type can still be identified although it is damaged, malformed or otherwise abberrant. Or even missing. As long as we have sufficient information in the original description or other publications, we can do it. And that is exactly my point. If we can positively identify the type as belonging to the species which name it bears, we can also positively identify any other fish using the same information. So who needs a type?

As for name changes, I don't see why having a type makes this any easier. It may be a good idea to preserve the original name for one of the species resulting from a split. You don't need a poll; the revisor just does it. If the split has to be reversed, the name can stand (again) for the re-united cluster. The other names will fall into synonymy.

The more I think about this, the more I realise that the only function of a type is a formal one. We must have types because the Code says so. I believe we can name species effectively without clinging that name to a specific specimen.
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 468
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: To be or not to (apart)?

Post by cichla » Sun Aug 25, 2013 5:34 am

It seems that we are talking about different things. :shock:

I thought of the concept of the 'type' and how it enter the Code (historical part) :? . Willem seems to speak about the 'type' as the bases for definitions and the standards of comparison. Rico is the advocate who is talking about 'how the type is used under the rules'.

Well, Rico, I agree 100% with you in all what you said about the type in accordance with the Code 8) . However, the type was treated (reduced) as the name bearer only (in a strict nomenclature sense) after the seminal paper by Simpson (1940). And even in the sixties and seventies there were discussion about it.
And you are much to optimistic, Rico. Just to quote Günther as a source gives an abridged version of the story. Regan, Günther & Co were great zoologists who recognize such things. Nevertheless, according to Mayr (1982: The Growth of Biological Thought: Diversity, Evolution and Inheritance) you may find in text books up to 1920 (at least) the type concept (also concerning species taxa names) as a typological concept.

I can follow your point Willem 8) . And I agree, too. Nevertheless, the function of a type is (how you said) a formal one, but also a proof (evidence) that there is something. If we do not have such type(s) everyone could name every 'fantasy' as a new species (see Nessie). What the majority does not know is that there is a published scientific name for Nessie. :lol:

Post Reply

Return to “Taxonomy”