Which clade should be named?

New cichlid species and taxonomy
Post Reply
User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Which clade should be named?

Post by cichla » Sun Sep 15, 2013 6:19 am

Some people like to name every single clade of a phylogenetic tree (based on molecular data) as a different species, even the members of these clades do not differ phenotypically (fig 1). In other cases, where clades are well differentiated phenotypically, but do not differ genetically, they are treated as one species (fig. 2).

So, what do you think? Which clades should be named? The morphologically (see fig 2) different (but not necessary genetically different) clades or the genetically distinct (see fig 1) but phenotypically not distinguishable clades.

Thanks
Attachments
alle_gleich.jpg
Fig. 1
alle_unterschied.jpg
Fig. 2

Bas Pels
Posts: 2247
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: Which clade should be named?

Post by Bas Pels » Sun Sep 15, 2013 7:23 am

Actually I would like to ask what gene / genes were used to draw the schedules.

Most often mitochondrial genes are used to establish a genetic distance, and I can only say that such a genetic distnce does not reflect much.

Mitochondria are inherited only from the mother, and the role of the mitochondria in the cell is to turn food into energy (more specifically, the citric acid cyclus and the NADH2 - ATP take place in the mitochondria). Therefore, mitochondria are an ideal tool to establish last contact, but any found difference is meaningless.

Meaningless, because this genetic difference does not reflect any adaptation or change in the building of the fish. After all, the morphology of any animal is determined by some of the genes in the nucleus, not by genes in the mitochondria

Su, assuming the genes were taken from mitochondria, the differences in fig 1 tell me we are talking about an old species, which most likely inhabits a few places which have not been in contact with each other recently (it could, for instance be an European river fish, the Rhine, the Meuse, the Elbe and the Thames used to be connected in the ice ages, flowing north west over the dry north sea plateau into the Atlantic ocean. But as the fishes have not changed their morphology, they are still the same species

If, however the used DNA comes from the nucleus (and is not involved in the production of some enzyme or so) I might have to reconsider the above.

In fig 2, the fish differ genetically and morphologically (I think the used colours reflect dissimilarity, not necessarily colour morphs) and are most likely different species.

But then, I think drawing up a ‘tree of life’ is the most scientific part, this can be done wrong or right. If I study one gene and someone else uses another one, ideally we get the same tree (better put, we best use both results in order to get closer to the right tree, the one we could easily verify if we had a time machine)

Drawing lines, however, where 1 species stops and another starts is not only science. People can draw other lines, have good arguments for both lines, and continue to disagree. Apparently more than just science plays a role here.

Bas Pels
Posts: 2247
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: Which clade should be named?

Post by Bas Pels » Sun Sep 15, 2013 7:29 am

If I may change the first picture a bit,
probleem.jpg
probleem.jpg (52.06 KiB) Viewed 5057 times
In this situation, I would assume that one population of the ols species needed to change a lot - while the other populations remained as they were.

This changed populatioon might very well be a new species, but the unchanged populations, which are now no longer a mponophyletic group (after all, the black fish in the middle is more closely related to the blue one than the other members of the black group) is, to me, still the original species.

User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: Which clade should be named?

Post by cichla » Sun Sep 22, 2013 1:03 pm

Thanks, Bas. Well, you are right, there are sequences which are more informative (to reconstruct the phylogeny) than other parts of the genome. However, I am just looking what we should name. This is why I like to simplify the issue.

Take two 'scenarios': [1] there are clades of genetically well differentiate cluster (Fig. 1) which are morphologically indistinguishable, and [2] there are clades which are genetically more or less identical (let the genetic distance below the level where separate species usually recognized) but differ in phenotypic appearance (Fig. 2).

So, just vote for case [1] or [2]. No sophiscated answer is necessary. I just like to know if you guys like to name phenotypically or genetically different clades.

User avatar
Darrell Ullisch
Posts: 371
Joined: Thu Aug 25, 2005 10:54 pm
Location: SW Michigan

Re: Which clade should be named?

Post by Darrell Ullisch » Tue Sep 24, 2013 3:06 pm

I would be more likely to name fish based on the genetic variance.

Experience tells me that morphology can be highly influenced by environment, the lack of genetic variance in that group suggests that the differences are from species elasticity more than genetic drift. Young from one population might actually grow up to look like another population if they are raised in the other environment (I have actually experienced this with certain killifish). Eventually one might see some genetic drift due to geographic separation or genetic adaptation to the varied environments, but at the point in time suggested by your clades it is premature to separate them as species. The behavioral differences would also be of interest, though they currently have no standing in nomenclature.

For the genetically separated clades, it is actually the opposite argument with the same factors. The fact that the fish are all morphologically similar is likely due to environmental similarity. However, probable geographic separation and sufficient time to allow genetic drift have caused them to be less compatible at the reproductive level, even though they are still using the same forms and possibly even the same behaviors. The key in my mind would be reproductive incompatiblility; it doesn't have to be complete, but should at least be limited to one or two generations.
There are two kinds of error: blind credulity and piecemeal criticism. Sound skepticism is the necessary condition for good discernment; but piecemeal criticism is an error. - Egyptian proverb

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 800
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: Which clade should be named?

Post by Willem Heijns » Sat Sep 28, 2013 3:38 am

Here is an illustration showing another approach to this issue. It leads us into the world of genotypes and phenotypes.

Image

Below the blue dotted line I have drawn nine different genotypes, which can be clustered into three groups, based on their relative differences. This raises the question: what is a genotype really? Is it a unique (individual) combination of DNA or is some variation within one genotype allowed? These questions also apply to phenotypes, by the way.
The genotype sets a constraint on the possible phenotypes that can be produced by it (a cichlid genotype could never produce a reptile phenotype). This is shown by the black arrows. Within this constraint, phenotypes are further shaped by the environment (i.e. natural selection). I have pictured three different scenarios. In scenario 1 the three (green) genotypes have produced only one phenotype, indicating genetic variation. In scenario 2 each (blue) genotype has produced one corresponding phenotype. Scenario 3 is an example of phenotypic plasticity. Two of the three (pink) genotypes have produced more than one phenotype. Now what does this mean for species delimitation? I have inserted the view of both lumpers (green arrows) and splitters (red arrows) into the picture.

If genotypes are leading, the three green genotypes constitute one species with one phenotype. The blue genotypes show both genetic and phenotypic variation within a single species. The same goes for the pink genotypes, a single species with even more phenotypic variation. This is of course according to the lumpers, who will distinguish three clusters of genotypes forming three species. Splitters would recognize nine species (one for each genotype). With the green genotypes, this causes a problem. Their phenotypes are indistinguishable.

The phenotypes tell a different story. If only phenotypes are studied, nine different phenotypes will be recognized and nine different species will be distinguished. Additional genetic research would then yield the structure shown in the picture. Scenario 1 still concerns one species (with intraspecific genetic variation). Scenario 2 may lead lumpers to the conclusion that the genetic differences are too small to warrant specific status. Thus one species with both genetic and phenotypic variation is recognized. Scenario 3 would probably puzzle them. Still, lumpers may conclude only one (but rather variable) species to be present in this scenario. Splitters would distinguish nine species, albeit not the same nine species as recognized in the genotype approach.

I hope this approach triggers the response this topic deserves. Any comments?
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: Which clade should be named?

Post by cichla » Sun Sep 29, 2013 1:50 am

Thank you, Darrel. Yes, the morphological appearance may be influenced by environmental conditions. On the other hand there are phenotypical characters in genetically similar populations which are independent of such environmental factors. Nevertheless I get your point of view, and your contribution is interesting and very welcome.
Many thanks to Willem. The figure is fantastic and very informative, well done!
I think the phentypically different forms in ''Scenario 2'' (in Willem's sketch) might be named as separate species.
Best, IS

Bas Pels
Posts: 2247
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: Which clade should be named?

Post by Bas Pels » Sun Sep 29, 2013 2:24 am

I think Willem made a little but significant error:
Willem Heijns wrote: Within this constraint, phenotypes are further shaped by the environment (i.e. natural selection).
While it is true phenotypes are influenced by the environment, this is not natural selection.

A fish which is, given the proper diet, is yellow, will not be yellow without this diet. Many snail eating cichlids only develop the jaws to eat these snails if fed said snails. Without them in the aquarium, they don't develop the strong jaws needed to do so, and will therefore be unable to crush snails.

This is just an explanation of the differences between phenotype and genotype.

Although it is true natural selection selects on phenotypes, it influences the genotype. After all, this selection decides who will breed - that is, will pass on its genes. And therefore, natural selection makes the genotype for the next generation.

Natural selection works slowly, over the generation.

There is a huge difference between a fish adapting and a species adapting to the environment. An individual cannot change its genes, so the adaptation is phenotypical a species does change its genes while adapting, so this adaptation is genotypical.

In my above contributions I started with asking what genes were studied. With good reason, as most often mitochondrial DNA - which does not dictate anything in an developing embryo - is used. Therefore one can wonder whether mitochondrial DNA is part of a fishes genotype. OK, if this DNA does not work, the fish will die, but that's all influence mitochondrial DNA has.

Nuclear DNA, which is for more complex, on the other hand does dictate what the egg will become. Nuclear DNA makes a fish into a fish. I would not be surprized if one could transplant fish mitochondria into a mammal - without any problems, where the mitochondria would continue to function - but I would not know how to do such an experiment. I'm not a scientist.

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 800
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: Which clade should be named?

Post by Willem Heijns » Sun Sep 29, 2013 9:55 am

Natural selection is the process that leads to differential reproductive success depending on how well organisms are adapted to their environment. As such, it acts upon phenotypes. Phenotypes with better adaptations will survive, whereas less adapted phenotypes will disappear. Natural selection thus directly influences the phenotypes we can observe. But where do these phenotypes come from in the first place?
DNA, the constituent of the genotype, continually changes through mutations. Many (most?) of these mutations don’t affect the phenotype at all. This has led scientists to wonder what the purpose of this particular mutated DNA might be. Therefore it is sometimes called “junk DNA”. For the mutations that do affect the phenotype, three possibilities present themselves:
1. the mutation is selectively neutral
2. the mutation is selected for
3. the mutation is selected against.
The second possibility will be the one prevailing, because organisms possessing the phenotype with this mutation will be reproductively more successful than others, leaving more offspring with the preferred phenotype. In summary: the (mutated) genotype ceates a phenotype which is favoured by natural selection. This is quite different from natural selection “making the genotype for the next generation”.

We must also make a clear distinction between natural selection and phenotypic plasticity. Phenotypic plasticty is the ability for an individual organism to adapt its phenotype to the environment. Cichlid teeth are a good example. Depending on the type of food present, a cichlid may develop molariform teeth or it may not. If snails are the (only) food present in their habitat, they may ontogenetically develop these molariform teeth to be able to crush the snail shells. If there is no such food restriction, the molariform teeth probably will not develop. This phenotypic plasticity can be observed in individual cichlids having developed these molariform teeth during their growth and other individual cichlids (of the same species) having developed another type of teeth. The phenotypic character of the species, apparently selected for by natural selection, is “the ability to develop different sets of teeth”. Not the teeth themselves.
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: Which clade should be named?

Post by cichla » Sun Sep 29, 2013 11:59 am

Dear Bas, Dear Willem, Dear all,
thank you for your contributions. Actually, it touches only slightly the original enquiry. ;-) Nevertheless your posts are not only welcome but informative and inspiring, for sure.

Well, I think you are talking about different things. Yes, there is phenotypical plasticity, and - more important - the discontinuous activity of genes (silent genes). So, I am pretty convinced that it does not need a long time to developing new 'species' (or – if you like – new phenotypes).

Bas, your point of view sounds more Lamarckian to me, and Willem sounds more like Darwinian. Anyhow, I am convinced that there are evolutionary units (usually called species) which are not genetically separate but – nevertheless – representing independent evolutionary lineage.

What do you like to name? Phenotypically different cluster or cluster of populations which are just differentiated by genetically distance (without phenotypically differences) ? :?

Bas Pels
Posts: 2247
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: Which clade should be named?

Post by Bas Pels » Sun Sep 29, 2013 2:51 pm

Dear Cichla

Although I know what Lamarckian evolution is - adaptations made by parents are given to their offspring - I do not see any Lamarckian evolution in my contribution.

perhaps this
There is a huge difference between a fish adapting and a species adapting to the environment. An individual cannot change its genes, so the adaptation is phenotypical a species does change its genes while adapting, so this adaptation is genotypical.
sounds Lamarckian, but it is not. I wanted to point out the differences between a species adapting and an individual. I did not intent to write - and I never wrote - that the individual adaptations are how the species changes.

How the genes change while adapting? mutations and selection. Or, if the needed genes are already around, selection of the good genes might do the trick

Mind you, if the good genes are not around, and the mutations do not help, the species will not adapt / but most likely go extinct

User avatar
Willem Heijns
CichlidRoom Expert
Posts: 800
Joined: Wed Jun 22, 2005 4:18 pm
Location: Stiphout, Netherlands

Re: Which clade should be named?

Post by Willem Heijns » Mon Sep 30, 2013 3:15 pm

Sorry for the distraction Ingo. :?

I believe we should name every species we distinguish. But if we distinguish species based on differences (either phenotypical or genetic or both) alone, we disregard any other character (behaviour, reproductive isolation), frequently used in species concepts. The observed differences can of course be presented as a phylogenetic tree, but that doesn't help us in deciding where the limtis of the included species are.

Kullander's (1999) phylogenetic species concept is based on diagnosability in combination with some form of parental pattern (i.e. reproduction). Taking into account that published diagnoses in original species descriptions have until now all been based on morphological characters (hence the indistinguishable phenotypes in my scenario 1), it seems that he prefers phenotypes over genotypes when it comes to species delimitation. I'm tempted to follow.
Slàinte mhath!

Uilleam

User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: Which clade should be named?

Post by cichla » Thu Oct 03, 2013 9:11 am

Thanks, Bas.
Well, I have to say that – at least for me – Lemarckian and Darwinian are equal. Means both hypotheses have advantages and disadvantages.
By the the way, the popular Darwinism ''just'' explains the phenotypical appearance but not the genetically diversity.

Dear Willem,
Willem Heijns wrote:Sorry for the distraction ...
No, problem. The posts are of some relevance ;-)
Thanks a lot, Willem, that you point out your view.
I think that we should name phenotypcal different population (only), instead of naming populations (or groups of populatins) which differs only by genetical differences. And I agree to some to a certain extent with Kullander's definition of a species taxon.

This is what Ernst Mayr (the Darwin of the 20th century) thought. To name clades of populations which differs only by differences in DNA (or other chemical stuff) makes no meaning.

I believe we should name every population (or group of populations) which we are able to distinguished.

What is your view?

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

Re: Which clade should be named?

Post by Rico Morgenstern » Sat Oct 05, 2013 2:22 am

I do not believe in 'cryptic species' distinguished by genetic characters only. In cases of sympatry, it is not detectable if the genetic differences do correlate with 'functional species', and in allopatry the differences may just reflect some genetic drift. Anyway, regardless if genetic, biochemical or other any other character complexes are employed - the problem remains always the same, namely to distinguish, which character states are really diagnostic for a species and which do just characterize the sample at hand or reflect intraspecific variation. In morphological/anatomical, behavioral or other phenotypic characters this is more easy to prove, provided the student is sufficiently familiar with the taxonomic group studied and with the distribution of the used characters within that group.

Bas Pels
Posts: 2247
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: Which clade should be named?

Post by Bas Pels » Sat Oct 05, 2013 3:04 am

Let's go back to what evoultion actually is. Mutations - and genetic recombination - make differences, on which the natural selection works.

If mutatins do not cause any difference in the phenotype, there is nothing to select on - and therefore nothing has changed.

I graduated in chemistry, and one would assume I would therefore feel much more at ease with DNA research than with less exact morphology. Still this is not the case. I feel a lot of unease when species are deliminated based on mitochondrial DNA or DNA sequenses which do not code. Because not coding changes will - by definition - not result in any change in the organism - fish or someting else.

Unfortunately, this is what commonly is used to do the searches.

User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: Which clade should be named?

Post by cichla » Tue Oct 22, 2013 7:55 am

Thank you Bas and all.
Bas Pels wrote:I would therefore feel much more at ease with DNA research than with less exact morphology.
Is it so, are genetically differences are really more valuable than the morphology ones? How many changes it is needed in biochemical groundplan to obtain a new ‘appearance’ in form and function? Or does it connect with proteins which, perhaps, switch on and off particular sequences?

Bas Pels
Posts: 2247
Joined: Fri Mar 03, 2006 9:17 am
Location: Nijmegen - the Netherlands

Re: Which clade should be named?

Post by Bas Pels » Tue Oct 22, 2013 10:07 am

cichla wrote:Thank you Bas and all.
Bas Pels wrote:I would therefore feel much more at ease with DNA research than with less exact morphology.
Is it so, are genetically differences are really more valuable than the morphology ones? How many changes it is needed in biochemical groundplan to obtain a new ‘appearance’ in form and function? Or does it connect with proteins which, perhaps, switch on and off particular sequences?
Cichla,

I think you quoted me a bit unlucky: the quote now tells something I do not agree with

on would assume a chemist would prefer a DNA based tree, but this chemist (meaning me) does not

Your questions are - more or less - why I hesitate following DNA evidence.

What does a difference in DNA actually mean? If the DNA is not coding, most likely very little. If the DNA is coding for a protein or so, it might still not be that imortant (in many a situation, more than one aminoacid can be put in a certain position in a protein, without much influence on its function).

But in the very few places where the DNA regulates growth, it would matter a lot. An awfull lot.

Human DNA differs some 2 % from Chimansee DNA. Our proteins are mostly the same, so the coding DNA is also vertually identical. Scholars assume we differ mostly in non coding DNA - which may not matter - and in regulating DNA.

For instance, expanding childhood will make a species able to have a more complicated life, as childhoud is, in all species of primates, known for it curiosity. Thus only one regulator might have done a lot of the larger knowledge humans have compared to chimps. But one might never be able to find this regulator in the DNA - and therefore is is averaged out doing searches after relationships

User avatar
cichla
CichlidRoom Expert
Posts: 467
Joined: Sun Jul 02, 2006 11:31 am
Location: Berlin, Deutschland

Re: Which clade should be named?

Post by cichla » Wed Oct 23, 2013 2:34 am

Sorry. Bas. And thank you for clarifying.

User avatar
Juan Artigas
Administrator
Posts: 1432
Joined: Tue Nov 25, 2003 7:37 pm
Location: San Luis Potosi, México
Contact:

Re: Which clade should be named?

Post by Juan Artigas » Fri Oct 25, 2013 11:45 am

Thank you for the wonderful thread and graphics I have enjoyed very much! I was watching Willem’s great graphic as a base for expressing my opinion.

In Willem’s group one the variability of the genotypes that produce a single phenotype cannot be too large, as otherwise it would not likely be possible to have just one clear phenotype. Even the most perfect evolutionary convergence should be able to give after a careful morphological examination, as far as I know. On the other hand, if the variability is due to an ancient species genetic variability, which may reflect a great genetic distance as shown by Ingo, most genes should be able to express as phenotypic traits under one condition or other (otherwise they would be considered junk DNA), in which case those genes are there for the species survival under unexpected environmental conditions. In Willem’s number one group, my opinion is that it should be just one species described.

In Willem’s group two, one genotype gives rise to one phenotype, they do not cross, so I don’t see the conflict there, I believe those should be thee species for lumpers and splitters alike. Probably closely related species in case of Ingo’s figure 2, but one species each. Each genotype is adapted to one species as an entity that can thrive and reproduce.

In Willem’s group three, each genotype gives rise to several phenotypes, I assume on an environmentally caused predetermined fashion. That would be three polymorphic species I believe, probably a species in the process of split in the proper conditions. I don’t see any cross there, so I believe there are just three species there, as the phenotypes produced should be interbreeding to keep the integrity of their genotype.
Juan Miguel Artigas
Editor

The Cichlid Room Companion
http://www.cichlidae.info

User avatar
Tachymarptis
CichlidRoom Expert
Posts: 42
Joined: Thu Jul 07, 2005 1:08 pm

Re: Which clade should be named?

Post by Tachymarptis » Sat Oct 26, 2013 1:05 am

Hi all,

This is an interesting discussion with practical implications, since current taxonomists often exaggerate the importance of genotypes in determining new species. Even in well-studied taxonomic groups such as primates (especially south-american ones), there is a trend to multiply the species number with poorly significant external features (if ever).

In my opinion, there should not be a pre-formatted lecture grid. Each case has to be studied according to the importance of expressed characteristics in speciation of the taxonomic group to which it belongs, especially when these characteristics are of major importance in sexual selection, leading to quick raise of new species (while others are not important at all and may give rise to polymorphism inside one species). Junk DNA, which is abundant in old, weakly evolving taxa, often distort results given by molecularists. Thus, you may have different species which are very close genetically in some cases, in others, there will be a great variability. This is the case for example in Tropheus vs mbunas or Victoria haplochromines. The first ones are more ancient and have a greater genotype variability while they are extremely homogeneous externally, yet, there are only few species.

And to summarize about phenotypes, they are of major importance in my opinion, because expressed genes are more important than hidden, spare ones, in speciation, but phenotypical plasticity must be tracked down and put aside since it has no signification, except if it eventually give raise to some selection and genetic drift, of course.

Post Reply

Return to “Taxonomy”