We invite scientists and educators to take part in the ongoing discussion about the scientific issues raised in Explore Evolution. If you would like to submit scientific comments or critiques about Explore Evolution, please send them in. We will review them and address them on these pages. For some particularly relevant comments or critiques, we will ask the author’s permission to publish their response, so please include your name, address, e-mail address and phone number with your submission to: furtherdebate@exploreevolution.com

« The "Fact" of Evolution | Main | Molecular Phylogeny and Phylogenetic Trees »

Malcolm Gordon and the Origin of Tetrapods

John Timmer accuses Explore Evolution of what he calls the "find a Ph.D." approach: "if you look hard enough, you can find someone with a PhD who will say anything." In this instance, Timmer disparages the minority viewpoint of UCLA biologist Malcolm Gordon (a tenured professor, actually), who has argued that the tetrapods may have evolved polyphyletically (i.e., more than once).

It's the textbook catechism again: why bother with citing some lone dissenter like Gordon? Timmer counts noses, and the sum determines what is worthy of attention. Claim that the scientists cited in EE pale in numbers to those who support the catechismal view, and voilá, case closed. There is no controversy and we can all go home.

This is science by census. But does Timmer really want us to believe that numbers of scientists, and not the evidence and how best to interpret it, is what matters?

As it happens--to play along with Timmer's counting-noses game--Gordon developed his view with the late UCLA paleontologist Everett Olson (a former president of the Society of Vertebrate Paleontology), in Invasions of the Land: The Transitions of Organisms from Aquatic to Terrestrial Life (1995), a book published by Columbia University Press. More recently, Gordon articulated his ideas with the Australian paleontologist John Long. But, as Timmer says, these are just another couple of Ph.Ds--you know: find a Ph.D, he'll say anything.

So let's look at the evidence. A review of the literature shows that there is much more to this story than Timmer lets on.

Surveying the problem of the overall picture of tetrapod evolution, Gordon (1999, 338) writes:

Despite the large volume of publication, however, the underlying reality remains unchanged: everything we know is circumstantial and indirect, and what actually occurred remains unknown.

This sentiment was later confirmed in part by Takezaki et al. (2004). They compared sequences of 44 nuclear genes encoding over 10,400 positions in their attempts to resolve the phylogenetic relationships between the coelacanth, lungfish and tetrapod lines. They write:

Apparently, the coelacanth, lungfish, and tetrapod lineages diverged within such a short time interval that at this level of analysis, their relationships appear to be an irresolvable trichotomy. (2004, 1512)

These findings amplify what Gordon (1999, 339) said five years earlier:

Thus there are significant variations regarding conclusions derived from molecular biological data sets, and differences between various parts of the morphological and molecular data sets.

Gordon goes on:

The living lungfishes and the coelacanth represent tiny, randomly selected remnants of ancient groups that were numerous, varied, and widely distributed in the Devonian. One can only wonder at how accurate, or even relevant, the relationships that we estimate to exist between these organisms today may be with respect to the actual phylogenetic relationships of their basal groups. (1999, 340)

Gordon's main point is that the biogeographic distribution of the tetrapods in the Late Devonian, coupled with the incongruence of molecular data, coupled with a knowledge of the range of environments occupied by early tetrapods, support the contention that the tetrapods may have arisen polyphyletically. The assumed sarcopterygian progenitors in the Late Devonian had low offspring dispersal ranges and limited geographic ranges, yet the early tetrapods they supposedly evolved into also occupied separate and limited geographic ranges, and had limited dispersal. Many of the earliest tetrapods inhabited environments from shallow marine tidal areas to brackish environments to fresh (Blieck et al., 2007).

However, these groups were also widely separated without any apparent environmental continuity between them at the time of their evolution. Late Devonian tetrapod species are "highly endemic" (Clack 2006, 184), meaning that they are "restricted to the locality or region where they have been collected" (Blieck et al. 2007, 229). The fossils come from sites many thousands of miles apart.

Thus, the phylogenetic series reconstructed in familiar evolutionary cladograms include taxa rarely found together as fossils. Cambridge University paleontologist Jennifer Clack, an expert on this evidence, notes that "taking the tetrapods sites worldwide, one thing is obvious: they lie scattered over the globe in places that were remote from each, on separate continents, even in the Devonian" (2002, 99). "These forms," note other paleontologists working on the puzzle (Zhu et al. 2002, 720), "seem to have achieved worldwide distribution and great taxonomic diversity within a relatively short time." This paleo-biogeographical puzzle raises significant evidential difficulties for monophyletic (single origin) scenarios.

Weighing these paleo-biogeographic challenges, Clack (2002, 99) considers the possibility of polyphyletic tetrapod origins, but then dismisses that hypothesis as less likely than the monophyletic scenario:

The alternative, that tetrapods radiated independently from lobe-fins that had originally been euryhaline [salt-tolerating] and subsequently lost their salt tolerance, seems even more unlikely and countered by the detailed similarities that are found in the tetrapods now known from over the world.

Here Gordon disagrees--and we have a case study in the fragility of the "consilience" of data lauded by Timmer.

Timmer argues that a "consilience" of different lines of evidence strongly favors the catechismal (monophyletic) tale, and faults EE for neglecting this consilience (e.g., the putatively mutually reinforcing molecular and anatomical data). He complains, for instance, that EE says nothing about the methods of cladistics, the approach within biological systematics that organizes taxa by shared characters: "A description of cladistic methods," he writes, "doesn't appear at all in EE."

But it is an open question whether molecules do reinforce morphology. Furthermore, as Gordon wryly observes (1999, 339)--and as is generally known among systematists--cladistic methods presuppose common ancestry:

First, since the analyses [of tetrapod relationships] were all done cladistically, the underlying phylogenetic model in all cases was monophyletic. A single "main line" of tetrapod evolution is assumed to have existed in all cases. Possible polyphyletic scenarios were methodologically and philosophically excluded as implausible.

The widely-used software packages that implement cladistic methods will try to arrange molecular and anatomical data (characters) into a monophyletic tree, come what may. Some of the characters will end up as homologies-- i.e., as similarities caused by common ancestry--and others as homoplasies--i.e., as similarities not caused by common ancestry--but the assumption that a monophyletic tree exists somewhere in the data is not up for grabs. Cladistic methods generate monophyletic trees, because they can't help but make such trees: that's what the methods were designed to do.

As Gordon's skepticism about cladistics indicates, behind the public proclamations that molecules confirm morphology, which Timmer recites, is an extensive scientific debate about the dangers of circularity in systematic methods. These questions are well-known to working systematists.

Could students hear about these questions? Why not? Is the catechism really more important?

EE concerns itself, therefore, with the logically prior question of "How do biologists infer (know) that all organisms, or some group of organisms, share a common ancestor?" That's a question students need to be able to answer, weighing the evidence pro and con, before they take up the merits of cladistics (which assumes the truth of monophyly as a first principle).

References Cited
Blieck, A., G. Clement, H. Blom, H. Lelievre, E. Luksevics, M. Streel, J. Thorez and G. C. Young. 2007. The biostratigraphical and palaeogeographical framework of the earliest diversification of tetrapods (Late Devonian). Geological Society, London, Special Publications volume 278. pp. 219-235.

Clack, Jennifer. 2002. Gaining Ground: The Orign and Evolution of Tetrapods. Bloomington, IN: Indiana University Press.

Clack, Jennifer A. 2006. "The emergence of early tetrapods." Palaeogeography, Palaeoclimatology, Palaeoecology 232:167--189.

Gordon, Malcolm S. 1999. "The Concept of Monophyly: A Speculative Essay." Biology and Philosophy 14:331--348.

Long, John A. and Malcolm S Gordon. 2004. "The greatest step in vertebrate history: a paleobiological review of the fish-tetrapod transition." Physiol. Biochem. Zool. 77:700-19.

Takezaki, Naoko, Felipe Figueroa, Zofia Zaleska-Rutczynska, Naoyuki Takahata and Jan Klein. 2004. "The Phylogenetic Relationship of Tetrapod, Coelacanth, and Lungfish Revealed by the Sequences of Forty-Four Nuclear Genes." Molecular Biology and Evolution 21:1512-1524.

Timmer, John. "A biologist reviews an evolution textbook from the ID camp." September 24, 2008. http://arstechnica.com/reviews/other/discovery-textbook-review.ars.

Zhu, Min, Per E. Ahlberg, Wenjin Zhao, and Liantao Jia 2002. "First Devonian tetrapod from Asia." Nature 420:760-1.


TrackBack URL for this entry:


This page contains a single entry from the blog posted on February 24, 2009 1:26 PM .

The previous post in this blog was The "Fact" of Evolution .

The next post in this blog is Molecular Phylogeny and Phylogenetic Trees .

For Discussion on the Debate