Palaeos Palaeos Phylogenetics
Systematics "Evolutionary Phylogeny"

Phylogenetics: "Evolutionary Phylogeny"

The need for a new synthesis

I grew up reading books on paleontology and evolution even as a kid in the 1960s. I always felt and still feel at home with the evolutionary systematics of the 50s and 60s. It was so easy, obvious, and intuitively common sense. This was before the cladistic revolution of the late 70s and early 80s. I remember reading about it when I was at University, and being puzzled by the whole thing. It was only much more recently, having gone online in the 1990s that I got a handle on it. More recently in researching for the upgrade of Palaeos I got a better handle on it all. Having thus spanned a major phylogenetic paradigm revolution, I have the advantage of being able to see both perspectives, and the strengths and weaknesses of each. Therefore I have no compulsion to accept one without question, and hence limit myself to a particular viewpoint. Moreover, it seems to me that every advance in science is made by including and adding new insights and methodologies to those that came before, not rejecting them. This is why I reject attempts to force evolutionary systematics into a quantifiable statistical mold, or to use cladistics to map phylogeny through deep time (rather than simply test different hypotheses of how the phylogenetic tree may be ordered). There is no reason why phylogenetic systematics cannot be added to the body of knowledge and insight of evolutionary systematics. It does not also follow that the insights of Hennig and Gauthier disprove or replace the insights of Mayr and Simpson, or vice-versa. A very different matter is incongruencies between Synapomorphy-based and Algorithm-based cladistics, or morphology and molecular based phylogenies. Hence the need for a new approach that includes all perspectives, insights, and methodologies: cladistic, evolutionary, morphological, molecular, developmental, systems approach, and even philosophical and artistic. Nothing less, I honestly believe, would be up to the task of mapping out the history of life on Earth. MAK111014

Mutual misunderstandings

I have noticed that there seems to be no love lost between cladists and evolutionary systematists (the latter being the "old school") . This is due to the misunderstanding of what cladistics is all about, i.e. both camps assume that cladistics is concerned with the same thing that evolutionary systems is, which is describing actual phylogenies, when it is not. It is simply a statistical method for evaluating different phylogenetic hypotheses. But because both supporters and opponents will often assume that cladograms are actual evolutionary trees with actual evolutionary phylogeny), these are claimed to be more up to date or modern than the bubble diagrams that evolutionary systematics uses. This is an example of cladistic literalism, the confusion of cladistic formalism (essential for hypothesis testing), which confuses cladograms with dendograms.

Because of this misunderstanding, cladistics has been criticised by many big name evolutionists, beginning with Mayr, (who coined the term cladistics for this school because he disliked it; ironically the name was then taken onboard by pattern cladists and became an established term), and continuing through to Richard Dawkins and Thomas Cavalier-Smith, and the authors of Res Botanica, support evolutionary taxonomy, and criticise cladistics among other things for excessive formalism and refusal to consider ancestral ("paraphyletic") groups, or its depreciation of the empirical approach of evolutionary classification as mere "intuition". Similarly, evolutionary systematics has been criticised by cladists for lack of easy repeatability (being based on imprecise, subjective, and complicated sets of rules that only scientists with experience working with their organisms were able to use) and for accepting paraphyletic taxa. That last is not a criticism but just a statement of different methodologies and interpretation. If it could be acknowledged that these two paradigms are not even talking about the same thing, perhaps there could be better relationships on sides. MAK111014

Among the controversial topics that are not given the attention they deserve by the majority of phylogenetics are Ancestral Characteristics, Supraspecific Taxa, Speciation, and Stratigraphic sequence (or Deep Time). MAK111018

Ancestors and Ancestral traits


Highlander - There can be only one! (Evolutionary systematics and Cladistics square off for the final showdown). Image from Media Unbound

Evolution is all about the common ancestor. It's not that (as in the standard linear model) humans evolved from (modern) apes, it's that humans and apes share a common ancestor. But one important difference between evolutionary systematics and cladistics is the way they each interpret this common ancestor in their diagrams and methodologies. When people look at a cladogram they assume that, as with an evolutionary systematic romerogram, the ancestors are the nodes. But cladistics is not intended to reconstruct the characteristics of actual deep time ancestors, or transitional forms, or missing or non-missing links. These are all paraphyletic taxa, the attachment to which strict cladistics is highly critical of. It is not that cladistics denies such organisms existed, but rather that, because it works on the species or individual level, it affirms that such taxa are unlikely to be fossilised. Archaeopteryx may be a close relative of the common ancestor of all birds, but it is not itself the actual common ancestor. Because the remains of the literal ancestor are not available, it is safer to compare the taxa we know of, and postulate hypothetical, ahistorical common ancestors as the nodes at the base of each monophyletic clade.

Most people, including scientists, prefer to think about evolution in terms of actual ancestors, not hypothetical abstractions. Richard Dawkins' criticism of the purely hypothesis-based cladistics and phenetics in his The Blind Watchmaker, might be due to a misunderstanding of cladistics in this regard. In any case, because evolutionary systematics usually deals with supra-specific ranks, it is not necessary to find the exact ancestor, as long as we have the general stem group, the bubble at the base of the bubble diagram. So even if Archaeopteryx lithographica Meyer, 1861 is unlikely to be the exact ancestor, another member of the family Archaeopterygidae or the order Archaeopterygiformes certainly would be.[1]

Whether or not actual (as opposed to hypothetic) ancestors are represented has important implications for the overall topology (shape) of an evolutionary tree because of the paraphyly rule. Evolutionary systematics, and authors like Simpson, Mayr, and Dawkins don't accept the paraphyly rule because the ancestral group (whether it be a species or supra-specific taxon) remains, as a taxon in space and deep time, unchanged, regardless of how many newly evolved lineages arise from it in the course of time. The (vertical) side-branch (budding) or daughter taxa (splitting, anagenesis) has no effect on the nature of the (horizontal) ancestors from which it arose. Early Triassic euparkeriids don't stop being small cursorial, armoured terrestrial predators just because some of them in the course of time tended to a more crurotarsal (crocodylian) lineage and others to a more ornithodiran (dinosaur, bird, and pterosaur) lineage. What paraphyly does is make vertical lineages (monophyletic branches) rather than groups of similar organisms the basic units. Application or nonapplication of the paraphyly rule can make a large difference to a taxonomic classification (Mayr & Bock 2002 p.181-2. Grant, 2003 p.1268). Again, this does not make one wrong or the other right. But each still claims "there can be only one". MAK111019

Stratigraphic sequence

Stratigraphic sequence, or the sequence by which fossils occur in geological strata, and hence the deep time chronological sequence of the organisms that they represent, is central to evolutionary systematics (since some of the founding figures there, G.G. Simpson, was a paleontologist). For the most part, this does not figure in cladistics or molecular phylogeny, although both these areas are concerned very much with the branching of clades in deep time (Hedges & Kumar 2009, see also molecular clock). Because the fossil record is known to be complete, the gaps are plugged with ghost lineages. One could say that cladistics puts parsimony ahead of stratigraphy, evolutionary systematics stratigraphy ahead of parsimony. This is why under Romer and Carroll the protothyrids are considered ancestral amniotes (the bubble at the bottom of the romerogram); they are primitive and they include the earliest known reptiles, whereas in cladistics they are considered a more derived or specialised off shoot, already diverged from the ancestral root and on on the road to diapsids. Of course, it may turn out that parsimony isn't everything here, or it is but new discoveries will overturn the current cladograms, and that the eureptilian condition is primitive for amniotes, just as the cryptodire condition (previously considered derived) now seems to be for turtles, but that's just baseless speculation at the moment (another way of looking at it is to understand the protothyrids as paraphyletic (Benton, 2004).

The obvious reason that neontology is emphasised is simply a pragmatic; extant organism prefer far more information than extinct ones. Although special preservation can reveal soft parts of fossil organisms, and although there has been some analysis of fossil genetic material which can be used in phylogenetic analysis, and even a project to sequence the Neanderthal genome, the simple fact remains that this material will never be as complete as that of living organisms. For this reason, cladistics emphasises crown groups (defined by extant, rather than fossil, organisms) as the basis for taxonomic entities.

The problem, from a deep time rather than a pragmatic perspective, with crown groups is that in the grand scheme of things this a purely relative and arbitrary category, as it depends on a particular period in geological or even historical time, and hence is always changing. Because the baseline is uniformly the present, there is almost an anthropocentric bias, because (pragmatic conditions aside), why should the present moment be any superior from an evolutionary, deep time perspective? Regarding the problems of basing crown groups on extant taxa see also Extinct or Extant. An alternative definition does not require all members of a crown group to be extant, only to have resulted from a "major cladogenesis event", but then there is the question of how this is to be defined.

Add to this conflicting definitions of groups like tetrapoda, are tetrapods defined morphologically (first appearance of limbs), phylogenetically (closer to land animals than lungfish or coelacanths), or in terms of crown groups (the common ancestor of all living (not fossil) tetrapods. The ideal of course is to include all perspectives. MAK111019

One field of phylogenetic analysis that does incorporate both stratigraphic sequence and morphological data is stratocladistics. This follows cladistic principles such as Bayesian logic and parsimony, but adds stratigraphic ordering. In this way, temporal data participate along with conventional character data in determining the most parsimonious hypotheses (Fisher, 1994, Fisher, 2008, Bodenbender & Fisher 2001). Yet after almost two decades stratocladistics is still very much a minority position, although one that deserves greater coverage.

The synthesis of stratigraphic paleontology and molecular phylogeny has been much more successful, with fossils used to determine the minimal age of branching points in the molecular tree of life, and thus calibrate the molecular clock (Donoghue & Benton 2007). The difference here is the premise that that molecules change at a constant rate, and if this can be determined using a few well defined calibration points (e.g. the divergence of synapsids and sauropsids) the whole tree of life can be dated in deep time. For this reason, molecular phylogenists have a greater interest in stratigraphy, where it applies to the branching points of major lineages, than cladists, who are concerned with finer scale divisions on the tree, and often have to deal with an incomplete fossil record and fragmentary material. MAK111020

Speciation and Dendrograms

Speciation is the process by which new species appear through natural evolution. This is a central theme of evolutionary systematics, which grew out of the so-called modern synthesis of Darwinian natural selection and Mendelian heredity. There seem to be at least three modes of speciation.

The Uncanny X Men

This is not how speciation works! Wikipedia, graphic copyright © 1941-2011 Marvel Characters, Inc.

Anagenesis is the transformation of one species into another over the course of geological time. Here the original species disappears and is totally replaced by the new species. This is how evolution is thought of in the popular imagination; along with implications of superiority of the new species (which itself ties back to old Great Chain of Being thinking). Hence the classic, and classically misinterpreted, March of Progress meme, and sci-fi and pop sci-fi images and tropes such as mutant superhero (in the X-men comics, Homo sapiens superior, the mutants with superpowers, and the new species emerging alongside ordinary human beings [2], and Transhumanism (posthumanity will succeed humanity). All of which conflates biological evolution with cosmological singularity. A more scientific, less mythopoetic, way of approaching this subject would be to look in the fossil record for instances of change or substitution of one species by another that is very similar to it, but occurs at a higher stratigraphic level. Here there are various taxa - foraminifera, ammonoids, trilobites, cenozoic mammals, with an excellent fossil record, and where many instances could doubtless be found. O'Keefe & Sander 1999 provide a case study of among mid Triassic pachypleurosaurs, and its interpretation using phenetic, cladistic, and stratigraphic methodologies.

Cladogenesis, also less imaginatively referred to as Splitting, involves the division (perhaps through geographic isolation or other factors) of an ancestral species or parental lineage into two or more daughter lineages or species, with the result that, like anagenesis, the ancestral species totally disappears. Only now, rather than one species replacing it, there are two or more. An example of this is allopatric speciation, whereby, e.g., a geographic barrier isolates population groups, results in the disappearance of the original species. Allopatric speciation, which occurs over long time dimensions, and it divides the ancestral species into more or less equal portions has been long advocated as the main speciation mechanism, especially in the zoological literature. (W. R. Elsberry via W.J. Hudson; Coyne & Orr, 2004, Horandl & Stuessy 2010, p.1643).

Finally, Budding, which like Splitting is a rather prosaic English term (in this case with nuances of gardening) rather than an interesting and scientific Greek one, is when a new species or population appears, (through the divergence of a small group of populations, without affecting the ancestral of the parental lineage in any way (Mayr & Bock 2002). As a result both the ancestral and the new species or stem group continues alongside each other (although perhaps geographically isolated). Obvious examples include cases of peripatric speciation after geographical isolation of a small group of populations. This is expected to happen mostly after colonizing events by a few individuals, then followed by rapid speciation and adaptation to new environments. Recent evidence from biogeographical studies on both animals and plants suggests that peripatric speciation may be more common than previously thought, since dispersal, even transoceanic dispersal, explains many disjunct distributional patterns. Buddings of this kind are often connected to a high amount of phenotypic change in the derivative species, which undergoes drift and adaptive change in the new ecological situation. In contrast, the source populations are neither in any novel environment, nor under any novel selective pressure." (Horandl & Stuessy 2010, p.1643-4)

Now, when Cladistics is presented simply as an empirical means of testing phylogenetic hypotheses, there is no problem. Mapping speciation is not necessary to cladistics and molecular phylogeny, indeed, it can be highly misleading, because because they are statistical (cladistics) or phenetic (molecular phylogeny) methods of formulating phylogenetic hypotheses, not historical accounts of one species transforming into or giving rise to another. It is only with the misinterpretation of cladograms as literal deep time phylogeny that things get awkward. This is because such a phylogeny could only recognise cladogenesis where the ancestral lineage divides into two (and only two) daughter species. However, something like allopatric speciation fits this model of symmetrical divergence, but this is no longer regarded as the predominant mode of speciation. (Rieseberg & Brouillet, 1994, Horandl & Stuessy 2010, p.1644 )

The following is a very interesting dendrogram (which is not a cladogram!) by Wang et al 1999, in which different modes of speciation are included, as well as deep time (stratigraphy) and chronospecies. This opens the way to a whole new way of doing phylogeny, which can incorporate both cladistics (phylogenetic systematics) and evolutionary systematics. To give it a fancy name I've informally coined it evolutionary phylogeny.

Phylogenetic systematics of the Borophaginae

Diagram by Wang et al 1999, p.339, showing the phylogenetic systematics of the Borophaginae, plotted against time (horizontal axis), and incorporating budding and chronospecies (anagenesis). The caption reads: "Stratigraphic distribution and postulated phyletic relationships for the Borophaginae. Although essentially based on our cladistic analyses, the phyletic relationships represent further speculations about the cladogenetic or anagenetic events, based on our assessment of their morphology and stratigraphy." From American Museum of Natural History Digital Library - Phylogenetic systematics of the Borophaginae, © American Museum of Natural History

A dendrogram such as the above one is however only possible where there is a fairly adequate fossil record. This works for Cenozoic mammals, some marine invertebrates, and even perhaps dinosaurs in a few narrow windows, such as the Morrison Jurassic and Latest (Campanian-Maastrichtian) Cretaceous. Sampson et al. 2010 for example provide a detailed cladogram, plotted against time (so technically a chronogram), of the Ceratopsinae of North America. Looking at their diagram it is not difficult to see where anagenesis might come in, especially where sister taxa occur in the same geographic locality and immediately succeed each other in time. Postulating such paraphyletic taxa does not contradict cladistics, because cladistics, as we often emphasise, is not about actual historical phylogenies. In this way, there is an opportunity for a reconciliation between cladistics and evolutionary systematics, each complimenting the other, so that there is no longer a need for there only to be one. MAK111019


[1] The reclassification of recent finds of archaeopterygids, like Jinfengopteryx and Anchiornis, as troodontids and the even more recent find of Xiaotingia, an ally of Anchiornis and Archaeopteryx, makes the position of the latter within Paraves far less clear cut. One of us (RFVS) thinks troodontids are derived archaeopterygids both because of the phylogenetic confusion and the arm length being the only distinguishing character. One thing we know with the relative certainty of the placement of Anchiornis within Archaeopterygidae and the uncertainty of the placement of the latter is that Anchiornis may be the earliest bird (Aves sensu Benton 2005) beating Archaeopteryx by 5 Ma. RFVS111203

[2] although technically this is budding because mutants and humans co-exist.

images not loading? | error messages? | broken links? | suggestions? | criticism?

contact us

content by MAK111019, edited RFVS111203, last modified MAK130316

Creative Commons License
Unless otherwise noted,
the material on this page may be used under the terms of a
Creative Commons License.