Reptile & Amphibian Forums

Hello folk,

reading the later dicussion between CKing and Richard Wells in the privious thread "Pogona" I would like to ask for comments on what makes a relieable taxonomic proposal that will be widely accepted?

A good point to start this discussion might be the following statements (taken from WW's page, http://biology.bangor.ac.uk/~bss166/update.htm):

"Taxonomy is a matter of personal opinion..." H.G. Cogger, 1985, quoted in Golay et al., 1993.

"Taxonomy is a matter of consensus..." Golay, 1993

"Taxonomy is a matter of evidence... " Wüster, 2002

Well, I think it obivously is a little bit from every thing. Personal opinion (base on knowledge), consensus (e.g. species concepts, etc.) and evidence (the way personal opionion, consensus and results are presented in the argumentation).

If taxonomy would only be a thing of personal opinion (based on knowledge), proposals, as those written by Hoser would have to be accepted, wouldn't they?
Consensus would acutally still accept propsals based on e.g. the BSC, wouldn't they?
Evidence seems to be the most convincing thing in taxonomy. But evidence also is a matter of knowledge and the objective view of the results and argumentation presented in a paper.

Which methods should be used to be evident? Some workers prefer morphological characters, others say that DNA analysis would be the most objective thing and therefore the most evident. But which one to use? mtDNA or nuclear DNA? Can every base pair really be used as a character? Is it a consensus today, not to use the BSC anymore because the reproductive isolation can not be tested in the wild? What about personal opinion towards evidence and consensus?

just to start a maybe interesting discussion...

Cheers,
Wulf
-----
http://www.leiopython.de - the white-lipped python site -
http://www.herpers-digest.com - herp related eBooks search -

To be widely accepted, a taxonomic proposal would probably need to have the following:

1. publicity. No proposal is going to achieve wide spread acceptance unless it is widely known.

2. data. No proposal is going to be accepted unless there is data in support of it.

3. controversy. If there is controversy over a taxnomic proposal, the chance of it being accepted is lessened.

4. authority. A taxonomic proposal is much more likely to be accepted if it is proposed by a well known specialist in the field and/or it is published in a prestigious scientific journal. Conversely, a taxonomic proposal is less likely to be accepted if it is published in a book, a herpetocultural magazine, an amateur herpetological journal (such as one of the regional herpetological society journals) and/or if it is published by, say, a student or an amateur in the field.

5. test of time. If a proposal (even if one that has garnered little support when published) has withstood the test of time, then it is likely to be widely accepted eventually.

6. ideology. A proposal based on ideology is most likely accepted by those who share the same ideology and rejected by those who have a different ideology. For example, taxonomists who split a species because of his/her subscription to the evolutionary species concept will generally find acceptance among adherents to the same concept but rejection by those who subscribe to a different species concept.

7. fashion. A proposal to split taxa, for example, on the basis of phenetic differences is much more likely to be acceptable to most biologists when phenetics was fashionable, as it was during the 1970's. A phenetic classification is much less likely to be accepted today, since cladistics is fashionable.

8. intangibles. Sometimes it is nearly impossible to figure out why one proposal is accepted and why another is rejected.

As to what one can do to influence acceptance of a taxonomic proposal, that is another topic altogether. One can, for instance, ask a friend who is an author of a field guide or checklist to include one's own proposal to give it instant credibility. If a taxonomic change is adopted by a field guide, many readers may believe that this proposal may have strong evidentiary support, or that it has withstood the test of time, or that it was a consensus among the specialists. If that is the case, a proposal that is controversial or unsupported by data may nevertheless become a "fait accompli," even if it may be undone entirely a decade later.

Hi CKing,

6. ideology. A proposal based on ideology is most likely accepted by those who share the same ideology and rejected by those who have a different ideology. For example, taxonomists who split a species because of his/her subscription to the evolutionary species concept will generally find acceptance among adherents to the same concept but rejection by those who subscribe to a different species concept.

That's exactly the point!
Ideologies always seem to ignore facts and knowledge. Things that were right once (and perhaps still are) are simply ignored or being argued.
Species concepts are not more than tools that may help to provide someone's findings. One concept perhaps does better than others, but does that mean that all others are wrong? It sometimes look like that.

7. fashion. A proposal to split taxa, for example, on the basis of phenetic differences is much more likely to be acceptable to most biologists when phenetics was fashionable, as it was during the 1970's. A phenetic classification is much less likely to be accepted today, since cladistics is fashionable.

Same here! Are phenetic characters nowadays inevident or do "modern" taxonomists only accept finding based on what they call "state of the art" methods?

Cheers,
Wulf
-----
http://www.leiopython.de - the white-lipped python site -
http://www.herpers-digest.com - herp related eBooks search -

>>Hi CKing,
>>
>>

>>6. ideology. A proposal based on ideology is most likely accepted by those who share the same ideology and rejected by those who have a different ideology. For example, taxonomists who split a species because of his/her subscription to the evolutionary species concept will generally find acceptance among adherents to the same concept but rejection by those who subscribe to a different species concept.
>>

>>
>>That's exactly the point!
>>Ideologies always seem to ignore facts and knowledge. Things that were right once (and perhaps still are) are simply ignored or being argued.

A lot of what is blamed in "ideology" is simply due to the fact that our ways of thinking about evolution have changed, partly as a result of the diveristy of different tools that we have available. When things like the BSC were being formulated in the mid 20th century, reproductive isolation and morphology were pretty much the only sources of evidence that were available to researchers. Today, we have access to a much greater variety of tools, which give us access to different types of information, particularly phylogenetic information. I would argue that it would be very sad and worrying if these new tools and new insights had NOT affected the way we view the natural world and classify it.

>>

>>7. fashion. A proposal to split taxa, for example, on the basis of phenetic differences is much more likely to be acceptable to most biologists when phenetics was fashionable, as it was during the 1970's. A phenetic classification is much less likely to be accepted today, since cladistics is fashionable.
>>

>>
>>Same here! Are phenetic characters nowadays inevident or do "modern" taxonomists only accept finding based on what they call "state of the art" methods?

Two points here:

First, phenetic methods continue to be employed for things like describing patterns of geographic variation, diagnosing species limits etc., which they do very well. Phenetic methods are rarely used above the species level because classification here is normally based on phylogeny, which phenetic approaches cannot recover.

Second, the phenetic vs. cladistic dispute is not about whether the methods are "state of the art" (as far as technique goes, both approaches can be as simple or as sophisticated as you like), but on the philosophical basis of what we are trying to reflect in classification. Cladists wpuld argue that since there is only one true phylogeny, we should attempt to reflect that in classification. This has become the most widely accepted view in systematics, whether one likes it or not.

Cheers,

Wolfgang
-----
WW Home

WW wrote:
'A lot of what is blamed in "ideology" is simply due to the fact that our ways of thinking about evolution have changed, partly as a result of the diveristy of different tools that we have available. When things like the BSC were being formulated in the mid 20th century, reproductive isolation and morphology were pretty much the only sources of evidence that were available to researchers. Today, we have access to a much greater variety of tools, which give us access to different types of information, particularly phylogenetic information. I would argue that it would be very sad and worrying if these new tools and new insights had NOT affected the way we view the natural world and classify it.'

Me:
I would agree that the availability of molecular techniques have revolutionalized systematics. I disagree that these new techniques have caused biologists to change their world views. The two most popular alternatives to the Biological Species Concept, namely the Phylogenetic Species Concept and the Evolutionary Species Concept, were both formulated decades before molecular techniques became available. The PSC, for example, was promulgated by Willi Hennig, whose entire methodology was formulated without the benefit of molecular techniques. In fact, when Hennig's methodology first became popular in the 1970’s, all of Hennig's followers were morphologists. Ironically, it is the availability of molecular techniques, which threatened the careers of many systematists who relied exclusively on morphology, which drove them to embrace Hennig's cladistic methodology in the first place. Even more amusing is that the cladistic methodology the morphologists embraced totally ignores morphological disparity when classifying organisms. It shows how desparate molecular techniques had made the morphologists.

Perversely, if the Hennigians have their way, there will be no need for morphologists since morphological disparity will no longer be considered when classifying organisms, and since molecular techniques have largely dominated systematics in recent years. By embracing cladistics, the morphologists have in fact made their own discipline irrelevant! Little wonder that the only discipline in which morphologists still rule is paleontology, in which molecular techniques are rarely applicable, since DNA data are rarely available from fossils.

WW wrote:
‘Cladists wpuld argue that since there is only one true phylogeny, we should attempt to reflect that in classification. This has become the most widely accepted view in systematics, whether one likes it or not.”

Me:
It is true that there is only one true phylogeny, but it is equally true that, as George Gaylord Simpson pointed out, phylogeny cannot be observed, and phylogeny has to be inferred, often from data that are not directly phylogenetic in nature. Since different systematists would most likely infer phylogeny differently, there is virtually no hope of achieving a consensus when one must classify organisms strictly on the basis of phylogeny. Paraphrasing Simpson, there is no known way of determining all the dichotomies (or other fractionations) in any phylogeny or their relative dates among fossils, making their classification both theoretically and practically impossible.

To be useful, classification must be stable. The sole purpose of the existence of the ICZN and its long lists of rules, for example, is the promotion of taxnomic stability. G.G. Simpson has, as Darwin and many Darwinians have done, expressed the view that to be stable and useful, a classification must take into account both phylogeny and morphological disparity. G.G. Simpson wrote:

“...I will state briefly what I believe to be both the consensus and the most sensible basic approach to classification. Classification cannot be a complete and detailed expression of phylogeny, but it should be consistent with a reasonable estimate of phylogeny. ...In a consistent classification, an inference will be that all the members of a taxon had a common ancestry, although, if the ancestry was a single species (it need not be) or even a single genus or family, that ancestry is not necessarily formally made a member of the descendant taxon. It is also a definitely established factor in phylogeny that some lineages and resulting taxa diverge more widely and evolve more rapidly than others. The degree of difference between taxa in their evolved or derived characters is therefore phylogenetic and is a useful criterion of categorical level. Classifications are artifacts constructed primarily for their usefulness in biological thought and communication. It is therefore desirable to be conservative regarding them and to change a classification radically only when it has in some way come clearly into conflict with the simple criteria stated in this paragraph.”

Reference
Simpson, George Gaylord 1983. Fossils and the history of life. Scientific American Books. W. H. Freeman and Company. New York San Francisco

Hi Wulf,

>>Consensus would acutally still accept propsals based on e.g. the BSC, wouldn't they?

Consensus is what develops over time after a proposal is first made in the literature - an increasing proportionof workers will either use or disregard the new arrangement. Also, where you look will make a difference. If a new taxonomic proposal is well supported, then academic publications will usually accept it quickly, whereas it takes a bit longer to percolate into the herpetocultural literature and when it comes to the taxonomic literature - consensus becomes synonymous with intertia. The average toxinological journal comes straight out of the 1960s when it comes to the taxonomy used.

>>Evidence seems to be the most convincing thing in taxonomy. But evidence also is a matter of knowledge and the objective view of the results and argumentation presented in a paper.

Evidence is primarily data, and second their analysis and interpretation.

>>
>>Which methods should be used to be evident?

That or those best suited to answering the question you are asking.

This requires in particular an awareness of the strengts, weaknesses and limitations of the different methods and markers.

>> Some workers prefer morphological characters, others say that DNA analysis would be the most objective thing and therefore the most evident.

Depends on your organisms and what question you are asking. If, for instance, you want to test whether two differentiated "forms" occur sympatrically where they meet, or whether there is an intergrade zone, a morphological analysis may be entirely adequate and much better than mtDNA.

On the other hand, it is certainly true that very often, the best bet is to use different markers in combination - molecular morphological markers together can be reciprocally illuminating and their combination may be much more powerful than the sum of the parts.

>> But which one to use? mtDNA or nuclear DNA?

Depends on what you are trying to do. Also, what nuclear DNA marker? A gene sequence? Microsats? AFLP? These all tell you different things, make different assumptions, and have different logistic requirements.

A phylogenetic analysis at higher taxonomic levels is best done using both mtDNA and nuclear gene sequences. On the other hand, if you are trying to determine species limits, then a combo of morphology mtDNA, or even microsats or AFLP (if you can get the required sample sizes together) would be much more appropriate.

> Can every base pair really be used as a character?

Depends on context. If you are running a phylogeny on one single gene or a single non-recombining set of genes, then yes. On the other hand, if you have 20 morphological characters and 1000 variable base pairs, then that would be a different question.

> Is it a consensus today, not to use the BSC anymore because the reproductive isolation can not be tested in the wild?

First, let's distinguish between species concept ("what is a species" and species diagnosis criteria ("how do we diagnose it?".

Under just about any evolutionary paradigm, a species is basically a separate evolutionary lineage. Most of the so-called "species concepts" such as the BSC are in reality centered around diagnostic criteria - the BSC simply states that reproductive compatibility is the criterion for recognising species. I think pretty much everyone would agree that reproductive isolation DOES indicate separate species status, since two lineages that cannot exchange genes are clearly evolving separately. However, lack of reproductive isolation is no loger viewed as precluding separate species status by most practicing systematists nowadays, partly because of the impossibility of relevant testing for allopatric taxa.

Cheers,

Wolfgang
-----
WW Home

Hi Wolfgang,

>>Consensus would acutally still accept propsals based on e.g. the BSC, wouldn't they?

You wrote:

Consensus is what develops over time after a proposal is first made in the literature - an increasing proportion of workers will either use or disregard the new arrangement. Also, where you look will make a difference. If a new taxonomic proposal is well supported, then academic publications will usually accept it quickly, whereas it takes a bit longer to percolate into the herpetocultural literature and when it comes to the taxonomic literature - consensus becomes synonymous with intertia. The average toxinological journal comes straight out of the 1960s when it comes to the taxonomy used.

Ok, so far so good, but what makes them disregard a given proposal? I understand that weak arguments or results provided in a proposal make other workers do so, but proposals with adequate data are also often discareded, because of more or less "ideological dissensions".
On the other hand evidence-free papers are widely accepted. I remember Stull (1935) placing the white-lipped python (L. albertisii Peters & Doria 1878) as subspecies of Liasis fuscus Peters 1873 without providing neither reasoning nor any evidence. This proposal was widely accepted (Loveridge 1948, de Haas 1950, Brongersma 1951, 1953 and 1956). And then Underwood & Stimson (1990) in their phylogenetic study placed all the indo-australian python species recognized at that time to a single genus Morelia, ignoring the obvious morphological differences. So, what is "consensus" then? Is it the collective loose of one's way in taxonomy and forgetting the things learned once?

> Is it a consensus today, not to use the BSC anymore because the reproductive isolation can not be tested in the wild?

Your wrote:

First, let's distinguish between species concept ("what is a species" and species diagnosis criteria ("how do we diagnose it?"

Under just about any evolutionary paradigm, a species is basically a separate evolutionary lineage. Most of the so-called "species concepts" such as the BSC are in reality centered around diagnostic criteria - the BSC simply states that reproductive compatibility is the criterion for recognising species. I think pretty much everyone would agree that reproductive isolation DOES indicate separate species status, since two lineages that cannot exchange genes are clearly evolving separately. However, lack of reproductive isolation is no loger viewed as precluding separate species status by most practicing systematists nowadays, partly because of the impossibility of relevant testing for allopatric taxa.

I do understand that it is impossible to test reproductive isolation of allopatric populations in the wild, and because of that, reproductive isolation is no longer being viewed as a creterion for speciation. Anyway, I tend to agree with Mayr (1996) that allopatric populations form incipient races and that "They may due in time aquire the needed isolating mechanisms to function as well separated species." (Mayr, 1996). So, if there is a reduced gene pool and no gen flow occurs with other popualtions, allopatric population will - under the evolutionary paradigma - also form seperate lineages, wouldn't they? It just a matter of time The phylogenetists (?) would now perhaps argue they are not, if they have the same ancestor, but as a matter of fact, we actually all have the same ancestor. Again its only a matter of the time you look back in your historical view
If I got that right, an ancestor in the phylogenetic meaning is the one species state before there is a split-up in any major character (that's at least what the phylograms try to tell when branching).
But by agreeing to Mayr's first statement, I also have to agree that "Owing to the gradualness of the process of speciation, every incipient species at one time in its cycle goes through subspecies stage." (Mayr, 1996). Am I wrong, or would it be so, that if one follows Mayr's statements, every allopatric population would at least be recognized at sub-generic level? It's only a matter of time until they will have to be recognized as full species?

just a few thoughts

Cheers,
Wulf
-----
http://www.leiopython.de - the white-lipped python site -
http://www.herpers-digest.com - herp related eBooks search -

>>

>>Consensus is what develops over time after a proposal is first made in the literature - an increasing proportion of workers will either use or disregard the new arrangement. Also, where you look will make a difference. If a new taxonomic proposal is well supported, then academic publications will usually accept it quickly, whereas it takes a bit longer to percolate into the herpetocultural literature and when it comes to the taxonomic literature - consensus becomes synonymous with intertia. The average toxinological journal comes straight out of the 1960s when it comes to the taxonomy used.
>>

>>
>>Ok, so far so good, but what makes them disregard a given proposal?

Who do you mean by "them"? If you mean toxinologists, it's because they are pig-ignorant and don't bother reading the literature.

If you mean other systematists, then there are avariety of reasons, usually to do with the data and/or interpretation thereof.

>>I understand that weak arguments or results provided in a proposal make other workers do so, but proposals with adequate data are also often discareded, because of more or less "ideological dissensions".

What you are calling "ideological dissensions" may have more to do with issues concerning the interpretation of data and the need (or otherwise) for the change. For instance, however good an mtDNA phylogeographic study is, it cannot truly elucidate the nature of contact zones without additional evidence, so many would be wary of proposals to split contiguous populations into species solely on the basis of an mtDNA phylogeny.

>>On the other hand evidence-free papers are widely accepted. I remember Stull (1935) placing the white-lipped python (L. albertisii Peters & Doria 1878) as subspecies of Liasis fuscus Peters 1873 without providing neither reasoning nor any evidence. This proposal was widely accepted (Loveridge 1948, de Haas 1950, Brongersma 1951, 1953 and 1956).

Sometimes simply because it was convenient at the time. Lumping was certainly in vogue in the middle 20th century, and the standards of evidence required then were not up to today's levels. If such a proposal were published today, then more evidenec would be required to make it "stick". The other question that has to be asked is what evidence existed at the time that albertisii was NOT conspecific with fuscus.

>> And then Underwood & Stimson (1990) in their phylogenetic study placed all the indo-australian python species recognized at that time to a single genus Morelia, ignoring the obvious morphological differences.

Classification is not about differences, it's about reflecting phylogeny. If the Indo-Australian pythons are monophyletic, then lumping them into one genus is one way of dealing with that result. Of course, later work by Kluge arrived at a very different conclusion regarding the phylogeny of the pythons.

>> So, what is "consensus" then? Is it the collective loose of one's way in taxonomy and forgetting the things learned once?

No. It is numbers of knowledgeable individuals agreeing on an issue on the basis of the evidence available. If there is little conclusive evidence, then you won't find consensus.

>>Your wrote:
>>

>>First, let's distinguish between species concept ("what is a species" and species diagnosis criteria ("how do we diagnose it?"
>>
>>Under just about any evolutionary paradigm, a species is basically a separate evolutionary lineage. Most of the so-called "species concepts" such as the BSC are in reality centered around diagnostic criteria - the BSC simply states that reproductive compatibility is the criterion for recognising species. I think pretty much everyone would agree that reproductive isolation DOES indicate separate species status, since two lineages that cannot exchange genes are clearly evolving separately. However, lack of reproductive isolation is no loger viewed as precluding separate species status by most practicing systematists nowadays, partly because of the impossibility of relevant testing for allopatric taxa.
>>

>>
>>I do understand that it is impossible to test reproductive isolation of allopatric populations in the wild, and because of that, reproductive isolation is no longer being viewed as a creterion for speciation. Anyway, I tend to agree with Mayr (1996) that allopatric populations form incipient races and that "They may due in time aquire the needed isolating mechanisms to function as well separated species." (Mayr, 1996).

They are functioning as separated species because they do not exchange genetic material. Isolating mechanisms are irrelevant to any discussion of species limits in forms that do not meet in nature.

>>But by agreeing to Mayr's first statement, I also have to agree that "Owing to the gradualness of the process of speciation, every incipient species at one time in its cycle goes through subspecies stage." (Mayr, 1996). Am I wrong, or would it be so, that if one follows Mayr's statements, every allopatric population would at least be recognized at sub-generic level? It's only a matter of time until they will have to be recognized as full species?

Whether subspecies should be recognised or not is a different kettle of fish altogether. Clearly, where you decide that two allopatric populations are different species or not is a difficult question. You could argue that two amphibian ponds separated by a motorway are in effect in different evolutionary trajectories, since geentic exchange will be interrupted by the near 100% risk of dying from Dunlop disease. However, few people would argue that they should be recognised as separate species.

The fact is that after geographic separation, populations will as a default option drift apart genetically, morphologically, and reproductively, but in no particular order. As they drift apart, the various criteria used for species diagnosis under different "species concepts" will start to be fulfilled but again, not in any particular order. So, for instance, you may have ecological and/or morphological differences evolving very rapidly if the two daughter species occur in different environments, but reproductive compatibility may persist much longer, especially since there will be no selection for isolation due to reduced hybrid fitness in allopatric populations. On the other hand, if the daghter populations continue to exist in identical habitats, morphological and ecological change may be minimal, but genetic changes (e.g. coalescence of mtDNA haplotype lineages) may occur.

By far the best discussion of species conceptsand species diagnosis is:

De Queiroz, K. (1998) The general lineage concept of species, species criteria, and the process of speciation. Pp. 57-75. In Howard, D.J. & Berlocher, S.H. (Eds.), Endless Forms. Species and Speciation. Oxford University Press, New York.

Email me if you want a PDF.

Cheers,

Wolfgang
-----
WW Home

WW wrote:
'What you are calling "ideological dissensions" may have more to do with issues concerning the interpretation of data and the need (or otherwise) for the change."

Me:
Precisely! Different schools of taxonomy will differ as to whether a particular set of data requires a change in the taxonomy. A Darwinian taxonomist, for example, has no problem recognizing paraphyletic taxa, such as Reptilia. A cladist, however, is intolerant of paraphyletic taxa. So taxonomists from these two different schools may treat the same data set quite differently.

WW:
'For instance, however good an mtDNA phylogeographic study is, it cannot truly elucidate the nature of contact zones without additional evidence, so many would be wary of proposals to split contiguous populations into species solely on the basis of an mtDNA phylogeny.'

Me:
I agree that mtDNA data cannot be used to delimit species. MtDNA is a more or less neutral character. Mutations in this molecule occurs randomly and at regular intervals. Hence it is a very good molecule to use as a molecular clock. If speciation were to occur at regular intervals like mutations in neutral molecules, then mtDNA would be very useful for predicting when speciation may have occurred in the past. Unfortunately, organismal evolution occurs independently of molecular evolution, a fact that even pioneer molecular systematists like Allan Wilson and his students and colleagues are fully aware of, hence it is not possible to equate mtDNA mutations with speciation events.

WW:
'Classification is not about differences, it's about reflecting phylogeny. If the Indo-Australian pythons are monophyletic, then lumping them into one genus is one way of dealing with that result. Of course, later work by Kluge arrived at a very different conclusion regarding the phylogeny of the pythons.'

Me:
That is of course the ideology of the cladists, who have decided to ignore morphological disparity when classifying organisms. Cladists only want to recognize holophyletic groups, which consist of one ancestor and all of the descendants of that ancestor. Holophyletic groups unfortunately tend to be heterogeneous, and they are not very useful for biologists who do not specialize in systematics. For example, reptiles mammals birds form a holophyletic group called the Amniota. However, this group has been split into three different taxa traditionally, namely Reptilia, Mammalia and Aves. Under the traditional arrangement, Reptilia is "paraphyletic" according to the cladists, since Reptilia do not consist of a single ancestor and all of the descendants of this ancestor. Some cladists have decided to eliminate paraphyly by including birds in their "Reptilia" while maintaining Mammalia as a separate group. Such a definition of "Reptilia" is of course more heterogeneous than the traditional usage, but it is the direct result of the kind of classificatory philosophy promulgated by WW and his cladistic colleagues. Both the Darwinians and the cladists have accused the others of recognizing groups that are not natural and are similar to groups that consist of "my car and three geese." Unless one is a cladist, it would be fair to say that it is clear which particular school is guilty of recognizing groups that consist of "my car and three geese." The cladistic definition of Reptilia (which includes birds) would appear to fit that description more than traditional Reptilia (which excludes birds). As a further example of the heterogeneity of cladistic taxa, simply recall Frost and Etheridge's "Chamaeleonidae", which combines the traditional Agamidae and Chameleonidae into a single family. Even more absurd is the proposal by some paleontologists that the whales and the hippopotamus be classified as a taxon called "Whippomorpha."

These examples do show that there exist major philosophical differences between different schools of systematists. There simply is no hope that these different schools will ever find a middle ground. Hence we are in for decades of taxonomic chaos to come.

WW:
"If a new taxonomic proposal is well supported, then academic publications will usually accept it quickly, whereas it takes a bit longer to percolate into the herpetocultural literature..."

Me:
Not so, many proposals are actually first adopted uncritically by the herpetocultural literature even though they remain controversial within academia. For example, Frost and Etheridge's arrangement of the iguanian lizards was adopted quickly by a field guide and the herpetocultural community even though many in academia prefer the old 3 family arrangement. Another example was the proposed change (on the basis of grammar) of the well known name Lampropeltis getulus to Lampropeltis getula. In both the literature (e.g. Keogh's morphological analysis of the American ratsnakes) and in personal communications with some academic herpetologists, the old name L. getulus continues to be used even though herpetocultural magazines took to the new name immediately.

WW wrote:
"Under just about any evolutionary paradigm, a species is basically a separate evolutionary lineage."

Me:
This is not so at all. According to Ernst Mayr, the most effective proponent of the BSC, a species is not a lineage.

WW:
"Most of the so-called 'species concepts' such as the BSC are in reality centered around diagnostic criteria - the BSC simply states that reproductive compatibility is the criterion for recognising species."

Me:
Again, I disagree. Organisms in nature behave in certain ways, and one of the ways they behave is to interbreed with only a relatively small number of closely related individuals. For example, birds do not interbreed with mammals, and fish do not interbreed with crocodiles, even though all of these organisms are part of the same lineage, namely Vertebrata. Further, even within a sublineage like birds or mammals, there is no interbreeding among most of the members of that sublineage. Hence it would be next to impossible to delimit a species if we choose to define species as lineages. Members of the same species do interbreed freely with each other but not with members of other species. This is what biologists observe in nature. This is how species maintain their uniqueness within a diverse biosphere. This is why the BSC is nearly universally accepted.

As verteran herpetologist H. M. Smith (1990, Herpetologica) explains:
"Refinement of the species concept in biology over the past century or so took its largest step toward fidelity to the facts of nature with the formulation denoted as the Biological Species Concept (BSC), advanced most effectively by Ernst Mayr in numerous books and articles."

WW:
"I think pretty much everyone would agree that reproductive isolation DOES indicate separate species status, since two lineages that cannot exchange genes are clearly evolving separately."

Me:
Indeed, reproductive isolation, paraphrasing Ernst Mayr, is the mechanism which is responsible for the causation and maintenance of discontinuities among contemporary living species.

WW:
"However, lack of reproductive isolation is no loger viewed as precluding separate species status by most practicing systematists nowadays, partly because of the impossibility of relevant testing for allopatric taxa."

Me:
Unfortunately many practicing systematists have gone off the deep end and they are espousing all sorts of ideological nonsense which bear no resemblance to how organisms behave in nature. Many practicing systematists have taken a huge leap backwards by embracing the typological species concepts that were once popular in the 19th century, before the modern synthesis showed how scientifically untenable typological thinking is. Unfortunately, science is not immune from fashion. Typological thinking, even though scientifically untenable, seems to be fashionable.

I wanted to take a slightly different tact, though I agree with many of the statements in the other two "clades" (couldn't resist) of this discussion.

To me what gains acceptance is quality of the work presented. I try to use multiple sources of data, I use morphology, morphometrics, electrophoresis and DNA sequencing. Where they disagree I consider this as something that must be reconsiled and I do not consider DNA sequencing any more objective than quality morphological analysis. I do not believe taxonomy can be done without morphology because if you want to do a phylogeny and don't include the described fossils then you are throwing out data. Thats not being objective, that's making a subjective decision.

I also hold to the ICZN and go by the ideal of "Refute it or Accept it". So if someone describes something, they only way to not accept it is to publish a refutation. Otherwise wether you like it or not it will be accepted, eventually, by many people. This is why Iverson, Georges and myself dealt with all the Wells and Wellington names for turtles one at a time in an Internationally published paper. So there could be no arguments, we either refuted them or accepted them. Hence I now accept the name Macrochelodina, even though many hate it.

I have had to deal with some very awkward taxonomic decisions, I did this by being objective, using the ICZN to guide the decisions and gathering as much data as I could.

Do all my taxonomic proposals go down well? No, and I could not care less, the data speaks for itself. I had one person come up to me once and say "I described one species in my whole career and you sank it"... Well it was not a species.

The point is, if the quality of the presented work is poor then the chances of it being accepted in the long term are poor. If the quality is good, then there will be no objections and it will be accepted.

So all up, good taxonomy will be accepted, do it right, publish it properly and there is no problem, no matter how people hate it or love it. Do it badly and those who hate it have all the ammunition they need to cause political rubbish for decades.

Cheers, Scott
Carettochelys.com

-----
Scott Thomson

If you believe you can or you can't you are always right.

Most systematists, even most molecular systematists, would agree that morphology cannot be ignored. The question is how one deals with conflicting data sets. Some people combine molecular data and morphological data and do a single analysis. Such an approach may seem objective, but if the data set is made up of mostly morphological characters, the signal from the molecular characters may be drowned out. Personally I do not favor such an approach. Like it or not, morphological characters, because they tend to be adaptive, are often less informative of phylogeny than neutral nucleotide substitutions. One need to look no further than Darwin's "Origin" to get a glimpse of how good systematists of the past try to search for good (informative) characters while avoiding the pitfall of bad (uninformative) characters. Darwin, for example, explained how rudimentary organs can often be the most informative of phylogeny. Unfortunately, many practicing systematists (especially the cladists and pheneticists who claim that their approaches are 'obejective') pay far less attention to character goodness than Darwin and his contemporaries and followers. Ignorance of character goodness is not objectivity.

I agree that molecular approaches are not bullet proof, but I believe they are better than morphological approaches if the systematists are careful about choosing outgroups and good characters. I have noticed a few recent molecular studies in which a member of the ingroup was chosen as the outgroup. For example, in a recent mtDNA study of the chorus frogs (Pseudacris, Hylidae), two members of the North American Hyla clade were chosen as the outgroup to this same clade. The result is an unreliable phylogeny that shows a number of anomalies. In another mtDNA study of relationships among Eurasian ratsnakes of the genus Elaphe, a derived descendant of Elaphe (namely Lampropeltis) was chosen as the outgroup of Elaphe! Again, the results from this study is anomalous as well. Choosing Lampropeltis as the outgroup is like picking Homo sapiens as the outgroup when one analyzes the relationship among the apes. Such silly mistakes often result in synapomorphies being mistaken as sympleisomorphs and vice versa. The result can be disastrous, as S. J. Gould pointed out on page 214-215 of his book "Wonderful Life." Symplesiomorphs that are mistakenly thought to be synapomorphies can result in the recognition of polyphyletic groups.

One way to detect silly mistakes made by molecular systematists is by using morphological data. Utiger et al., in a recent analysis, concluded that the American species Senticolis triaspis is a member of the ratsnake clade (genus Elaphe 'sensu lato') even though there is poor statistical support for that node. Morphologically, Senticolis lacks the intrapulmonary bronchus that is possessed by all American members and some Eurasian members of the ratsnake clade. Morphologically, prior studies also fail to establish a close relationship between Senticolis and any other colubrid snake. Without corroboration, Utiger et al.'s analysis could have been uncritically accepted.

It appears that most systematists do agree that morphological data cannot be ignored when attempting to construct a phylogeny. Systematists, however, may disagree strongly on how and how much morphological data should be incorporated.

Unfortunately, there are also a large number of systematists, namely the cladists, who will totally ignore morphological data once they have constructed a phylogeny. To them, the classification is the branching diagram. To them, morphological disparity has no role in ranking organisms. It is quite ironic that many cladists may use morphological data almost exclusively when constructing a phylogeny and then totally ignore the data when trying to classify organisms. It is the cladistic practice of ignoring morphological disparity when ranking organisms that the Darwinians and I find most objectionable.

I largely agree with what you say, you have read a little too far into the point I was trying to get across. Which was that good quality taxonomic datasets include various sources of information, which can be tested against each other. A series of internal checks so to speak. These internal checks make it clear for all to see that lots of data was cross-correlated to produce the result. Rather than some data, totally unsubstantiated and largely subjective.

You also raised this point and I too have used morphology to delineate species that really lumped under just using allozymes or mtDNA.

My specialty as a taxonomist/palaeontologist is morphology, anatomy etc. I am not a geneticist, but I utilise the work of my co-authors extensively, that is I work with people who add this to my work. I have nothing against genetics. I just do not see it as completely objective. Any more so than morphological data is if its done properly.

I know you are not fond of cladistics but what you seem to be referring to is what I call an arm-chair taxonomist. If you have never seen anything but the blood of the species you are working with there is not much you know. But thats personal opinion.

Cheers, Scott
Carettochelys.com

-----
Scott Thomson

If you believe you can or you can't you are always right.

Not sure what you mean by "completely objective." Molecular data is of course not infallible. I cited several instances in which the choice of outgroup really affected the reliability of the results. That said, molecules do offer something morphology has been unable to accomplish and may never be able to do so. For example, it would perhaps be impossible to answer the question of exactly how closely related the Neandertals are to living human populations. That answer finally came when mtDNA data became available for living human populations and from a piece of Neandertal bone.

Further, until recently, it is nearly impossible to know where a speceis such as Lampropeltis zonata or Charina bottae had originated. mtDNA shows that both originated in southern California and migrated north. Even more revealing is molecular data on the western skinks. Long considered to consist of but 2 morphologically distinct species (Eumeces skiltonianus and E. gilberti), it turns out that the gilberti morph has arisen independently from different populations of E. skiltonianus, even though the different gilberti morphs may interbreed freely with one another while being reproductively isolated from the parental skiltonianus. It would have been nearly impossible to reveal such complexity without molecular data.

You have also mentioned the lumping of species using allozyme and mtDNA data. I think that practice is untenable. Species consist of sets of actual or potentially interbreeding populations, and no mtDNA data can possibly tell us anything about reproductive isolation. To this date, we still do not know whether Neandertals are reproductively isolated from modern Homo sapiens or not. Hence the question of whether Neandertals and modern humans are the same species remain an open question.

There is no doubt about it, the study of phylogeny is best done with molecules, simply because there are so many more molecular characters than morphological ones and because many molecules are selectively neutral whereas morphological characters tend to be adaptive and therefore more likely to be the result of convergent evolution. Nevertheless, phylogeny per se really gives us an incomplete picture.

I have taken part of an actual branching diagram from a published paper. This tree is constructed from mtDNA data. Can anyone tell me how many species are represented in this tree? I seriously doubt that anyone can figure out the answer since mtDNA data can only give us information about lineages, and lineages are not species. The result will be revealed in a future post. Have fun in the mean time.

Hi folks,

I'm just testing my knowledge

CKing worte:

There is no doubt about it, the study of phylogeny is best done with molecules, simply because there are so many more molecular characters than morphological ones and because many molecules are selectively neutral whereas morphological characters tend to be adaptive and therefore more likely to be the result of convergent evolution. Nevertheless, phylogeny per se really gives us an incomplete picture.

Well, I agree that there are much more molecular characters than morphological ones, but as you already stated above "Molecular data is of course not infallible" and they might lead to wrong phylogenetic hypotheses.

In his PhD. thesis, B.N. Campbell (1997) examined the molecular phylogenetics of Boidae using the mitochondiral cytochrome b gene. He examined 47 blood samples as well as 143 tissue samples representing 44 species. As a result, Campbell concluded that data from cytochrome b gene were unable to resolve the phylogeny of the Pythonidae in most aspects.

Taggart et al. (2001) also revealed the incongruence between mtDNA and nuclear-DNA data in the genus Bothriechis and concluded that "...the mtDNA results may not reflect specimens relationship..." (Taggart et al., 2001).

As pointed out by Henderson (1997) mtDNA of West Indian taxa of Corallus varies only slightly from Corallus hortulanus, but there is a notable divergence between them in morphologic characters.

So, I would conclude that phylogenetic analysis only on the base of molecular analysis can lead to wrong hyptheses about the phylogeny of species.

There also seems to be still a general methodological problem in molecular systematics. The weighting of data as preferred by some taxonomists may influence phylogenetic relationships. "Subjective" weighting or rejecting of data seems to be more like a "...if they woun't fit, I'll make them fit..." thing and there are several methods to handle data sets used in phylogenetic studies.

Cheers,
Wulf
-----
http://www.leiopython.de - the white-lipped python site -
http://www.herpers-digest.com - herp related eBooks search -

Not sure what you mean by "completely objective." [/quote

Well basically that is me getting my back up against geneticists who claim that molecular analysis is the one, the only completelyobjective means to analyse phylogeny. But enough of that.

One of my biggest gripes against the way molecular analysis if often done is sampling presumed clades. They take a single specimen of a single species from a presumed species group and sequence one gene (and if its mtDNA its not really even that) and make a pronouncement on how revolutionary there findings on the relationships of the different clades are. Ridiculous, their analytical method does not meet any statistical challenge.

[quote]
You have also mentioned the lumping of species using allozyme and mtDNA data. I think that practice is untenable. Species consist of sets of actual or potentially interbreeding populations, and no mtDNA data can possibly tell us anything about reproductive isolation.

Yeah I probably should have put it the way Wulf did and say that the analysis failed to resolve the phylogeny. That was 3mtDNA genes and 1 Nuclear Gene by the way.

There is no doubt about it, the study of phylogeny is best done with molecules, simply because there are so many more molecular characters than morphological ones and because many molecules are selectively neutral whereas morphological characters tend to be adaptive and therefore more likely to be the result of convergent evolution. Nevertheless, phylogeny per se really gives us an incomplete picture.

Really, I have seen no evidence of that. Many molecular based trees are outright ridiculous, others are very good. Some tell us very important things leading to valid reassessment of the phylogeny including the morphology. There is no evidence that a given set of molecular sequences are neutral, they could be adaptive, neutral or random. Knowing which is which is like throwing darts at a dart board with a blindfold on. The molecular trees hope that enough of the molecules are neutral to override those that are deliterous to the tree. There is no way to know this, it is an assumption and this is why Molecular data is based on subjective analysis.

There also seems to be still a general methodological problem in molecular systematics. The weighting of data as preferred by some taxonomists may influence phylogenetic relationships. "Subjective" weighting or rejecting of data seems to be more like a "...if they woun't fit, I'll make them fit..." thing and there are several methods to handle data sets used in phylogenetic studies.

I agree, as I have outlined above. There seems to be a definite belief that they are outside the boundaries of statistically valid evidence. I harp on that one, yes, but I did minor in statistics. There are huge assumptions in the weighing of data because there is not necessarily any idea of what the genes in question do, so how can you weigh them. I do on occasion weigh morphological data but at least I can use the function of the character to guide the weighing. It too is not perfect and contains assumptions but I test this by getting the identical tree with weighed and unweighed data, the only differences is nodal support.

I use molecular data extensively, it is not all bad, but it is one tool. That is all, there are others which include morphology, and statistical analysis. Morphology is effected, visibly, by behaviour, so understanding the ecology and ethology of the species is also important. Good phylogenies require the use of as many tools as possible. Not one.

Cheers, Scott
Carettochelys.com

-----
Scott Thomson

If you believe you can or you can't you are always right.

If a molecular systematist were to presume that a genus or species group is a clade, then he/she really has blind faith concerning the soundness of prior morphological anlayses, the very analyses that he/she is testing/challenging using molecular techniques! Isn't it ironic?

Yes, molecules can be adaptive or neutral. Usually neutral characters evolve much more quickly than adaptive ones. It is often not difficult to tell whether a molecule is neutral or adaptive. For example, there is no reason to believe that SINE characters or microsatellites are adaptive. These characters have been used very successfully to study phylogenetic history. SINE characters, for example, are practically free of homoplasy. In molecular systematics, as it is in morphological analysis, character goodness and character choice are very important determinants of success or failure. The goodness of morphological characters, on the other hand, are often much more difficult to evaluate. For example, if a species of aquatic newt has smaller eyes than 3 closely related but more terrestrial species, is it possible to know whether the large eyes of the terrestrial species are the result of a single speciation event or 2 or three independent speciation events? In other words, is the large eyes of the terrestrial species a homology or a homoplasy? Sometimes it is almost impossible to tell without molecular data.

There is little doubt that the future of phylogenetic analysis will be based primarily on molecules. However, in the field of paleontology, in which molecular data is extremely difficult to obtain, morphological characters would still need to be used. Luckily, even though molecular evidence may be lacking, fossil data, specifically stratigraphic data, is not. These can be used to corroborate morphological analyses.
SINE characters

Not having read BN Campbell's Ph. D. thesis, I won't be able to comment on why his analysis failed to resolve relationships. In some cases, the inability to resolve relationships is caused by adaptive radiation, in which taxa diverge very quickly from each other, leaving little time for molecular changes to accumulate. In the case of North American Hyla, this is apparently the case. Several species form an unresolved polytomy when immunological technique was used, suggesting that these species evolved quickly from a single common ancestor and diverged quickly into many different species, ostensibly because of an absence of competition. Hylid frogs originated in the southern hemisphere (Australia and South America) and they did not enter North America until after the collision of the North and South American continents. When a species very similar morphologically to Hyla eximia entered North America, it quickly spread to the entire continent and even to Europe, while diverging into the various species of Hyla, Pseudacris, Acris, and Limaoedus.

The situation in Corallus is not unusal. Similar phenomena are well known even to Darwin. Darwin knows that there are vastly different rates of evolution in different lineages. Darwin has no knowledge of molecular evolution of course, but he and his colleagues are fully aware that the rates of morphological evolution differ widely among different groups of organisms and among different species within the same group. Species which have changed little over vast stretches of geologic time are often called "living fossils." Even within the human lineage, scientists know that the rates of evolutionary change among the apes is much slower than the rate of evolution of the lineage which led to Homo sapiens. Very early on in the development of molecular techniques, systematists realize that the rates of evolution of molecules are very different from the rates of evolution of morphological characters and that the two occur independently of each other. If you compare the mtDNA between human and chimp, the difference is smaller than the mtDNA difference between gorilla and chimp, but the morphological differences are much greater between human and chimp than between gorilla and chimp.

I agree with you that "...phylogenetic analysis only on the base of molecular analysis can lead to wrong hyptheses about the phylogeny of species." The same can also be true of morphological data.

I also agree that weighting can lead to a change in the phylogenetic hypothesis. Weighting is subjective and therefore it depends on how well the researcher understands the molecules with which he/she is working and how well he/she weights his/her characters. By the way, when a researcher treats all characters equally, he/she is also being subjective, since he/she is weighting all characters equally. As for the rejection of data, it can be subjective or objective. If somebody made a mistake in the analysis, such as picking a member of the ingroup as an outgroup, then there is little to do but to reject the data as being unreliable. Such rejection can be objective. If somebody rejects data simply because it contradicts his/her own preconceptions, then it is probably subjective.

>>In his PhD. thesis, B.N. Campbell (1997) examined the molecular phylogenetics of Boidae using the mitochondiral cytochrome b gene. He examined 47 blood samples as well as 143 tissue samples representing 44 species. As a result, Campbell concluded that data from cytochrome b gene were unable to resolve the phylogeny of the Pythonidae in most aspects.

Noone says moleuclar markers are infalible. For a start, as in all cases, you need the right tool for the right job, and you need to be lucky enough to be studying an organism where there is a story

Disclaimer: I have not seen Campbell's thesis. However, if, as I suspect, the different Pythonine lineages are very old, then a fairly fast-evolving gene like cytb may not be the best choice of marker due to heavy saturation. Second, it is also possible that the Pythonines are the result of a rapid adaptive radiation, i.e., the different lineage splits happened in quick succession. In that case, this is going to be difficult to resolve in any case.

>>As pointed out by Henderson (1997) mtDNA of West Indian taxa of Corallus varies only slightly from Corallus hortulanus, but there is a notable divergence between them in morphologic characters.
>>
>>So, I would conclude that phylogenetic analysis only on the base of molecular analysis can lead to wrong hyptheses about the phylogeny of species.

Referring specifically to the Henderson study (and by implication to a large number of offshore island species in general), then we must not confuse different parts of the story being told by different markers and the notion that "molecular analysis can lead to wrong hyptheses about the phylogeny of species". Chances are that Henderson's mtDNA phylogeny is spot on: the differentiated island populations are recent derivatives of mainland populations, but which diverged rapidly after getting to the islands - a very common story. The mtDNA information is not misleading in any way, and neither is the morphological data - they simply tell a different part of the story. Of course, you and the others are correct in saying that having both sources of evidence at hand greatly strengthens our ability to get to the bottom of the entire story.

The question of what one does with recently diverged populations like that is a subject of debate, and really down to personal philosophy/preference. Some like to recognise them as separate phylogenetic/evolutionary species due to morphological differentiation and allopatry, others would rather sink them back into the mainland species based on phylogenetic considerations.

>>
>>There also seems to be still a general methodological problem in molecular systematics. The weighting of data as preferred by some taxonomists may influence phylogenetic relationships. "Subjective" weighting or rejecting of data seems to be more like a "...if they woun't fit, I'll make them fit..." thing and there are several methods to handle data sets used in phylogenetic studies.

That is a bit of an overstatement. By and large, weighting methods in molecular systematics seek to exploit our understanding of how DNA sequences evolve to compensate for artefacts like transition bias etc. Moreover, these weighting approaches have now been supplanted by more rigorous model-based approaches such as Maximum Likelihood and Bayesian methods.

Cheers,

Wolfgang
-----
WW Home

WW wrote:
"Chances are that Henderson's mtDNA phylogeny is spot on: the differentiated island populations are recent derivatives of mainland populations, but which diverged rapidly after getting to the islands - a very common story."

Me:
Sometimes organisms do diverge rapidly when they are stranded on an island. However, rapid divergence does not always occur among populations that are geographically isolated. The California Tiger salamander (Ambystoma tigrinum californiense) for example, is geographically isolated from other subspecies of tiger salamanders and yet it is morphologically very similar to the other populations elsewhere. Reproductive isolation has not evolved, and in fact the courtship behaviors of A. t. californiense and A. t. tigrinum are identical. Introduced tiger salamanders hybridize freely with native tiger salamanders within the range of A. t. calliforniense. Conversely, the lack of geographic isolation has not prevented rapid speciation among the cichlid fishes in the African lakes. As S.J. Gould has often pointed out, evolution cannot be a slow process because environmental changes are not slow. Organisms which change too slowly will not be able to adapt and survive. So, whether rapid divergence will or will not occur really depends on how different the environment on the island may be and whether there are unfilled niches on an island. The lack of mammalian predators on some Indonesian islands has apparently allowed the Komodo dragon to evolve in isolation to fill these niches. Similarly, island tortoises tend to evolve gigantism in the absence of mammalian herbivores.

WW:
"The question of what one does with recently diverged populations like that is a subject of debate, and really down to personal philosophy/preference. Some like to recognise them as separate phylogenetic/evolutionary species due to morphological differentiation and allopatry, others would rather sink them back into the mainland species based on phylogenetic considerations."

Me:
As I have said elsewhere, typological thinking is fashionable. The so-called phylogenetic species and evolutionary species concepts are little more than typological species concepts in disguise. To many taxonomists, species are morphotypes, and therefore when a geographically isolated population has evolved a slightly different morphotype, it is often treated by the typologist as a separate species. Typological thinking is in fact so fashionable that even molecular systematists have adopted it. Some molecular systematists have recognized different mtDNA haplotypes as different species. After these mtDNA "species" have been delimited, the taxonomist then searches for morphological characters with which to define these "species." This was in fact the case in the naming of the southern rubber boa as a different species, although the morphological characters that supposedly define "Charina umbratica" are unreliable. Other systematists have also attempted to delimit genera using mtDNA haplotypes, for example Utiger et al., although they include hemipenial morphology in their analysis. The result is a big mess: a large number of erected/resurrected genera that cannot be defined morphologically nor can they be distinguished from each other morphologically. In fact, Utiger et al. did not even try to define the genera they recognize primarily on the basis of mtDNA haplotypes. These cases of taxonomic chaos could have been avoided if systematists take into accoutn morphological disparity in classifications and if they simply accept paraphyletic taxa as the inevitable result of the process of evolution.