Reptile & Amphibian Forums

The attached diagram has been redrawn from Duellman 2001, the Hylid Frogs of Middle America. Duellman based this diagram on da Silva's phylogeny. Notice the position of Hyla eximia relative to the Hyla versicolor group (which includes Hyla chrysoscelis). In both of the trees of Moriarty and Cannatella (2004), H. chyrsoscelis is basal to Hyla eximia. In Duellman's tree, Hyla eximia is basal to Hyla chrysoscelis. Moriarty and Cannatella's Pseudacris is based on cutting the tree at position B, thus including Hyla crucifer, Hyla cadaverina and Hyla regilla in Pseudacris, making it a heterogeneous genus with some secondarily terrestrial treefrogs with reduced toe pads and also some species traditionally included in Hyla because of their well developed toe pads. An alternative way for delimiting Pseudacris is to apply the cut at position A, thus leaving Hyla crucifer, Hyla cadaverina and Hyla regilla out of Pseudacris. Pseudacris, without these species of Hyla, is much better defined and more homogeneous, since all members are secondarily terrestrial hylids with reduced toe pads.

That group of secondarily terrestrial, toepadless species (if it's a clade as the tree shows; I don't know from New World hylids) seems like a good unit to carry the name Pseudacris. No argument from me.

But there's no need for that to render Hyla paraphyletic, or to split Hyla up into a million indistinguishable genera to avoid paraphyly. Just use the CLADE name Pseudacris, and recite after me: all Pseudacris are Hyla, but not all Hyla are Pseudacris.

It's called INCLUSION. Descendants never leave their ancestral taxa, they stay connected to the tree at the same point in history where they first budded off. That's why there is never a reason to recognise paraphyletic taxa (apart from species) when the tree is known. Just throw away the rigidity of Linnean rank, and there's no problem....
-----
John D. Scanlon
Riversleigh Fossil Centre
Outback at Isa
Mount Isa, Queensland, Australia

of course, since doing away with Linnaean ranks is a "newfangled" idea, CKing is bound to oppose it . . .

or will he simply oppose it on the basis that you and I accept and that we are both anti-Mayrian Cladists . . . LOL

Troy

I prefer to throw away the rigidity of useless Hennigian dogma rather than the useful Linnaean system. So do a majority of zoologists world wide. Gauthier and deQueiroz have been advocating for years that biologists do away with the Linnaean ranks, but their suggestion has not been widely accepted.

Paraphrasing R. F. Inger of the Field Museum, "[the Linnaean system], despite its limitations and despite our occasional faulty applications, is a [system] that has proved useful and I think we will continue to use it" (Mayr 1982).

Reference
Mayr, E. 1982. Of What Use Are Subspecies? Auk 99:593-595

`That group of secondarily terrestrial, toepadless species (if it's a clade as the tree shows; I don't know from New World hylids) seems like a good unit to carry the name Pseudacris. No argument from me.

But there's no need for that to render Hyla paraphyletic, or to split Hyla up into a million indistinguishable genera to avoid paraphyly. Just use the CLADE name Pseudacris, and recite after me: all Pseudacris are Hyla, but not all Hyla are Pseudacris.'

Doesn't that just render your nomenclature unnecessarily confusing? WTF do you call triseriata, for instance? Hyla triseriata? Pseudacris triseriata? Hyla Pseudacris triseriata?

`It's called INCLUSION. Descendants never leave their ancestral taxa, they stay connected to the tree at the same point in history where they first budded off. That's why there is never a reason to recognise paraphyletic taxa (apart from species) when the tree is known. Just throw away the rigidity of Linnean rank, and there's no problem....'

OTOH, if you know the tree you don't need the names to reflect phylogeny--you've got the phylogeny right there in front of you.

Patrick Alexander

Well yes, but natural objects that are likely to come up in conversation (e.g. species, or clades with distinctive apomorphies) might as well have names, otherwise communication becomes a bit laboured. That's why we have taxonomy at all.

To qualify my point about giving up higher ranks, that's not necessary in this case because one could use the widely accepted rank of subgenus for Pseudacris. But what you usually see when people use subgenera is that they feel compelled (under the Linnean straitjacket) to assign every member of the 'split' genus to one subgenus or another, which usually creates at least one more paraphyletic (i.e. imaginary) group. Same with 'species groups'; how often do you see people defining a bunch of monotypic 'species groups' in a large genus? Even at this informal level of classification, the Linnean habit makes people do strange and pointless things, and then medieval-minded pedants have something else to argue about when they've settled the angels-on-pins question (or applied radical epistemic doubt to questions of their own and each others' eye colour). If we stick to naming groups that have some evidence for monophyly, there's no need for redundant ranks and we can get back to discussing biology and evolution.
-----
John D. Scanlon
Riversleigh Fossil Centre
Outback at Isa
Mount Isa, Queensland, Australia

You claim that "species and or clades with distinctive apomorphies" are "natural objects." You also claim that species can be paraphyletic. Then you claim that paraphyletic taxa are "imaginary." How can a species, which is both paraphyletic and natural, be imaginary?

I agree with you that taxonomy is important for communication both among scientists and between scientists and the general public. A term such as reptile is universally understood among both scientists and the general public. By redefining reptiles to include birds, some taxonomists have made communication more difficult. Further, since mammals are also descended from a reptile, the exclusion of the mammals from the Reptilia creates a paraphyletic basal group of primitive amniotes, which were once considered reptiles but is not a paraphyletic group without a name. This paraphyletic group is still part of Amniota, but it now has no name, making it impossible for scientists who adhere to this sort of cladistic absurdity to communicate with each other. This is madness, not progress.

If clades are natural, how come we can no longer classify some groups (such as the paraphyletic basal amniotes) if we recognize other groups such as mammals and Gauthier's "Reptilia"?

CKing:
You claim that "species and or clades with distinctive apomorphies" are "natural objects." You also claim that species can be paraphyletic. Then you claim that paraphyletic taxa are "imaginary." How can a species, which is both paraphyletic and natural, be imaginary?

Me:
Species CAN be paraphyletic because bits of them can bud off as new species which are then more closely related to some members than others in the 'parent' species. BUT THEY GET OVER IT - because of reproductive continuity, a species can re-establish monophyly over time (as all individuals living at some later time will share common ancestry since the last speciation). Species are the most inclusive groups that can be 'ancestral to themselves' in this way (under the BSC). And THAT's why rules for naming species have to be different from those for naming higher taxa. Paraphyly in species is a passing phase; paraphyly in higher taxa is irremediable except by the taxonomic action of re-including the other descendants of the common ancestor. Paraphyletic higher taxa are unnatural because, in reality, those other descendants NEVER LEFT THE LINEAGE.

CKing:
I agree with you that taxonomy is important for communication both among scientists and between scientists and the general public. A term such as reptile is universally understood among both scientists and the general public. By redefining reptiles to include birds, some taxonomists have made communication more difficult. Further, since mammals are also descended from a reptile, the exclusion of the mammals from the Reptilia creates a paraphyletic basal group of primitive amniotes, which were once considered reptiles but is not a paraphyletic group without a name. This paraphyletic group is still part of Amniota, but it now has no name, making it impossible for scientists who adhere to this sort of cladistic absurdity to communicate with each other. This is madness, not progress.

If clades are natural, how come we can no longer classify some groups (such as the paraphyletic basal amniotes) if we recognize other groups such as mammals and Gauthier's "Reptilia"?

Me:
The clade called Reptilia is a useful thing to name because it is a natural object in the world. I notice you seem to have no trouble with the concept of Amniota, which is equally natural and also does NOT correspond to a unitary 'concept' in the languages used by the biologically naive or otherwise pre-cladistic general public (people like Romer whose concept of 'reptiles' might be restated as: amniotes that aren't mammals or birds; as well as those who group reptiles and amphibians as a single concept like Linnaeus' 'Amphibia'). If the problem is that Reptilia (unlike Amniota) has the same name as a traditional paraphyletic concept, that is something that can be clarified in a moment if somebody gets confused in the course of discussion. (Two different words can have the same origin and spelling, that's just one of the ways language evolves, like gene duplication.)

Cladists recognising a monophyletic Reptilia DOES NOT create a paraphyletic group of basal amniotes; that's what people like your heroes do. Organisms that are are amniotes but not reptiles or mammals form a paraphyletic assemblage that doesn't deserve a collective name in formal taxonomy, although we can easily refer to them collectively - as I just did. Of course they're all extinct, so the details of their biology are the province of specialists who can give an explicit list of the monophyletic or species-level taxa they are discussing rather than trying to adapt a folk taxonomy that just doesn't apply. Your objection is refuted.
-----
John D. Scanlon
Riversleigh Fossil Centre
Outback at Isa
Mount Isa, Queensland, Australia

Yes, indeed the basal amniotes are extinct, but if Gauthier is incorrect, i.e. if turtles are descended from an amniote that is more basal to the ancestors of the mammals, then some member of this group of extinct basal amniotes has left descendants but this group nevertheless remains nameless. Gauthier's "Reptilia" would also be polyphyletic or at best paraphyletic if this is the case. I think there is recognition of this fact as some have proposed classifying turtles in their own taxon. That is indeed quite a mess. Gauthier's Reptilia is also heterogeneous, as it includes reptiles which lack endothermic homeothermy and descendants of reptiles that do have a different sort of physiology than most reptiles.

Since you are a paleontologist, you do need to communicate with other paleontologists regarding fossil taxa. Currently, there is no generally agreed informal term for the paraphyletic basal amniotes among cladists so that everyone knows what everyone else is talking about. Under the rigid rules of cladism such a group cannot even have a formal name. They can be called non-mammalian, non-reptilian amniotes, but this is awkward and it is inaccurate since these early amniotes are, as far as one can tell, very much reptilian in their morphology and physiology. Under the system Romer and Simpson used, and the same system that is being used by most biologists, all amniotes can be classified formally in their own Linnaean taxon. The cladists therefore cannot eliminate paraphyletic groups. They simply put these taxa into drawers without any labels. The new system that the cladists are proposing is therefore not progress, but a step away from utility.

`Paraphyletic higher taxa are unnatural because, in reality, those other descendants NEVER LEFT THE LINEAGE.'

I think you're equating taxa with lineages. I also think that this is a mistake.

Patrick Alexander

I appreciate you responding to paraphyly in species vs. paraphyly in larger taxonomic groups. You made a good point: the lack of genetic reciprocal exclucivity in many species is "temporary", and lineage sorting leads to exclusivity as both species continue on their evolutionary trajectories.

johnscanlon wrote
"Species CAN be paraphyletic because bits of them can bud off as new species which are then more closely related to some members than others in the 'parent' species."

Higher taxa also originate by the process of budding. In fact, it is extremely unlikely that splitting, in which an entire higher taxon becomes extinct when a new higher taxon evolves, is the norm in evolutionary history. A small genus with a few species may eventually become extinct, but large genera such as Hyla, Rana, and Bufo can persist for long periods of geological time and remain paraphyletic, since they are morphologically conservative and since morphologically divergent taxa have budded off of them. The appearance of Pseudacirs, for example, did not, and almost certainly cannot, cause the extinction of the speciose genus Hyla. Hence the cladistic practice of disqualifying paraphyletic higher taxa is, as Mayr pointed out, "impractical, destructive and scientifically untenable."

LOL - what an accurate description!

Troy

`Well yes, but natural objects that are likely to come up in conversation (e.g. species, or clades with distinctive apomorphies) might as well have names, otherwise communication becomes a bit laboured. That's why we have taxonomy at all. '

Exactly--and that's why paraphyletic taxa are often the best solution.

`To qualify my point about giving up higher ranks, that's not necessary in this case because one could use the widely accepted rank of subgenus for Pseudacris. But what you usually see when people use subgenera is that they feel compelled (under the Linnean straitjacket) to assign every member of the 'split' genus to one subgenus or another, which usually creates at least one more paraphyletic (i.e. imaginary) group.'

And there's another problem solved by using paraphyletic taxa.

`Same with 'species groups'; how often do you see people defining a bunch of monotypic 'species groups' in a large genus?'

Actually, I'm not sure I've ever seen that.

`Even at this informal level of classification, the Linnean habit makes people do strange and pointless things, and then medieval-minded pedants have something else to argue about when they've settled the angels-on-pins question (or applied radical epistemic doubt to questions of their own and each others' eye colour). If we stick to naming groups that have some evidence for monophyly, there's no need for redundant ranks and we can get back to discussing biology and evolution.'

From my experience, if we stop worrying about monophyletic taxa we can get back to biology and evolution. Giving up on a comprehensible hierarchical naming system, on the other hand, seems like it'd just be pointlessly confusing.

Patrick Alexander

Patrick Alexander wrote (quoting me; numbers added):

`Well yes, but natural objects that are likely to come up in conversation (e.g. species, or clades with distinctive apomorphies) might as well have names, otherwise communication becomes a bit laboured. That's why we have taxonomy at all. '

(1) Exactly--and that's why paraphyletic taxa are often the best solution.

`To qualify my point about giving up higher ranks, that's not necessary in this case because one could use the widely accepted rank of subgenus for Pseudacris. But what you usually see when people use subgenera is that they feel compelled (under the Linnean straitjacket) to assign every member of the 'split' genus to one subgenus or another, which usually creates at least one more paraphyletic (i.e. imaginary) group.'

(2) And there's another problem solved by using paraphyletic taxa.

`Same with 'species groups'; how often do you see people defining a bunch of monotypic 'species groups' in a large genus?'

(3) Actually, I'm not sure I've ever seen that.

`Even at this informal level of classification, the Linnean habit makes people do strange and pointless things, and then medieval-minded pedants have something else to argue about when they've settled the angels-on-pins question (or applied radical epistemic doubt to questions of their own and each others' eye colour). If we stick to naming groups that have some evidence for monophyly, there's no need for redundant ranks and we can get back to discussing biology and evolution.'

(4) From my experience, if we stop worrying about monophyletic taxa we can get back to biology and evolution. Giving up on a comprehensible hierarchical naming system, on the other hand, seems like it'd just be pointlessly confusing.

My replies:
(1-2) What advantage does paraphyly have here? It always obscures actual relationships when these are known, and it is purely arbitrary what parts of the clade are excluded.

(3) I try to avoid groups with unwieldy numbers of species per genus, but the Australian skink genera Ctenotus and Lerista are examples where a species-group classification has been used including various monotypic 'groups' (e.g. in work by Storr where phylogeny was not a consideration).

(4) Cladistics IS a comprehensible hierarchical naming system; or what part of monophyly don't you comprehend? Recognising paraphyletic taxa involves arbitrarily and subjectively jacking up subgroups of a clade to a higher rank, adding 'information' of dubious value and deleting true information about their genealogical relationships. Just because it has been done by default since before Darwin was born, doesn't mean it makes any sense, or is intuitively more satisfying than monophyly. You may find it so, but I don't.
-----
John D. Scanlon
Riversleigh Fossil Centre
Outback at Isa
Mount Isa, Queensland, Australia

johnscanlon wrote:
(1-2) What advantage does paraphyly have here? It always obscures actual relationships when these are known, and it is purely arbitrary what parts of the clade are excluded.

Me:
Traditional taxa are not defined arbitrarily. They are recognized on the basis of both morphological disparity and phylogenetic relationships. Contrast this with recent taxonomic proposals by the likes of Utiger et al., who split the morphologically conservative genus Elaphe into nearly a dozen genera on the basis of mtDNA data. Their taxa are arbitrarily delimited since none of the genera they recognize can be morphologically distinguished from one another or from Elaphe. The reason they split Elaphe is because a number of morphologically distinct genera (e.g. Pituophis, Lampropeltis, Cemophora et al.) have budded off of Elaphe, rendering it paraphyletic.

The advantages of paraphyletic taxa are obvious:
1. There is no need to recognize contrived taxa due to excessive splitting
2. There is no need to lump taxa excessively, which is the only other alternative available to the cladists
3. Paraphyletic taxa can be more morphologically homogeneous than holophyletic taxa.

johnscanlon:
(3) I try to avoid groups with unwieldy numbers of species per genus, but the Australian skink genera Ctenotus and Lerista are examples where a species-group classification has been used including various monotypic 'groups' (e.g. in work by Storr where phylogeny was not a consideration).

Me:
If a group of species closely resemble each other and their common ancestor morphologically, then it is not wise to split it into many different genera, even if distinct lineages can be discerned (e.g. using mtDNA) and even though such a group may be paraphyletic. The resultant mess is far worse because there would then be many contrived genera that are morphologically indistinguishable from one another. Because they are placed in different genera, their close relationship is also obscured. The cladistic alternative, which is excessive lumping, creates an unwieldly large genus, which is also morphologically heterogeneous.

Johnscanlon:
(4) Cladistics IS a comprehensible hierarchical naming system; or what part of monophyly don't you comprehend? Recognising paraphyletic taxa involves arbitrarily and subjectively jacking up subgroups of a clade to a higher rank, adding 'information' of dubious value and deleting true information about their genealogical relationships. Just because it has been done by default since before Darwin was born, doesn't mean it makes any sense, or is intuitively more satisfying than monophyly. You may find it so, but I don't.

Me:
I am sure Patrick comprehends “monophyly”. In reality, the destruction of paraphyletic taxa involves arbitrarily and subjectively jacking up subgroups of a clade to a higher rank. Species which are classified by most systematists in the paraphyletic genus Elaphe are being subjectively and arbitrarily elevated into new genera by cladists who cannot tell us the difference between the genera they recognize. Splitting Elaphe into a number of morphologically indistinguishable genera such as Euprepiophis, Pseudelaphe and Pantherophis actually deletes information about the genealogical relationships among them and adds no new information whatsoever. Darwin, who understands evolution perhaps better than Hennig and most cladists, fully recognizes the utility of paraphyletic groups, for he recognizes that different lineages may have evolved morphologically at vastly different rates, and it would be impossible to codify these different rates of evolutionary change without recognizing paraphyletic groups.

Darwin wrote:
“I believe that the arrangement of the groups within each class, in due subordination and relation to the other groups, must be strictly genealogical in order to be natural; but that the amount of difference in the several branches or groups, though allied in the same degree in blood to their common progenitor, may differ greatly, being due to the different degrees of modification which they have undergone; and this is expressed by the forms being ranked under different genera, families, sections, or orders.

Darwin elaborated further:
“The forms descended from A, now broken up into two or three families, constitute a distinct order from those descended from I, also broken up into two families. Nor can the existing species, descended from A, be ranked in the same genus with the parent A; or those from I, with the parent I. But the existing genus F14 may be supposed to have been but slightly modified; and it will then rank with the parent-genus F; just as some few still living organic beings belong to Silurian genera.”

Me:
Darwin thus shows us that in order to codify the morphological changes the descendants of A have accumulated, these descendants should be removed from the same genus, and even the same family, as A. On the other hand, since the descendant of F has not changed, it should be classified in the same family and genus as its ancestor F. Paraphyletic groups are therefore not arbitrarily defined; paraphyletic taxa are the result of the need to codify evolutionary change. If there is no evolutionary change, then there is nothing to codify and evolutionary biologists do not have to recognize anything but holophyletic groups. But a world in which no evolutionary changes occur is an uninteresting one for biologists, and of course systematics would not even exist. Neither would human beings. In a world in which nothing ever changes evolutionarily, all we would find are blue-green bacteria.