Darwinius massillae, continued…

ResearchBlogging.orgI found a new paper in my reader this morning from the crew who published the first description and taxonomic statements about Darwinius massillae, Phillip Gingerich and his colleagues.  This paper is a reply to Williams et al. (2010), which called into question a lot of the conclusions about the place of Darwinius in the greater scheme of primate evolution.  For a quick re-cap of where we are in the debate today, you can read this, this, this, or Brian Switek’s paper here.

They open their newest paper with a statement that everyone can agree on:

A forty-seven-million-year-old primate Darwinius masillae from the middle Eocene of Messel in Germany is worthy of attention because it is one of the most complete fossil primates found to date.

That, it certainly is. Beautiful, too. Ida helps us paint a picture of what life was like for primates in the Eocene.

For many paleontologists, that would be good enough.  But other paleontologists live and breathe in order to make taxonomic statements (ie, This is related to That).  Consider these two species of paleontologists the Morphologist and the Cladist.

The Cladist needs to rely on the Morphologist, though.  When the Cladist wants to make a phylogenetic tree or cladogram, he or she needs to compile a list of character states, or traits, that are going to be used.  An important point that Gingerich et al make in this paper is that those traits have to be independent traits- developmentally, functionally, and evolutionarily.  If they are not, you are going to have a lot of redundancies in your phylogenetic tree, and you may end up showing that a certain group is more closely related than it actually is. One developmental character shift can affect many different aspects of anatomy.  A good Morphologist will understand the origins of different anatomical traits, and should be able to tell which traits are meaningful, and which are redundant.

Gingerich et al. re-state the traits which they found in Darwinius massilae which they interpreted as derived toward the Haplorrhine clade, as a result of shared ancestry.  Let’s review:

  • A cranium with a short rostrum (Or a head with a short nose).  This trait evolved in Haplorrhines after lemurs and lorises split off to form their own branch. Darwinius doesn’t simply have a short snout because it is a young individual, because size of the premolars, as well as the number of them, is reduced.  However, there are strepsirrhines with “short” snouts, and Haplorrhines with “long” snouts.  My main question is, when does a snout go from “longish short” to “short,” or “shortish long” to “long”?  Maybe “most” haplorrhines have a short snout and “most” strepsirrhines have a long snout, but does that make it a reliable trait?
  • A Mandible with a Deep ramus (or a “tall” jawbone).  The mandible became “deep” after tarsiers split from the rest of the haplorrhine clade, and Darwinius shares this trait.  Again, there are strepsirrhines with deep rami and haplorrhines with shallow rami.
  • A fused mandible.  Most mammals have a mandible that exists in two halves, a right and a left.  Anthropoid primates (monkeys and apes) have a fused mandible, so that it exists as one solid entity.  Darwinius has a “partially” fused mandible, which is fused on the bottom (the ventral side) and open at the top (the dorsal side, where the teeth are).  This trait is also a “generalization with exceptions.”  However, the exceptions are spread out throughout the primate tree, which indicates that this trait probably evolves very “easily.”  It may even be related to loading during life, so that it’s not a trait that is under selection and can pop up in any primate who subjects its jaws to heavy stresses.
  • Spatulate incisors (Spoon-shaped incisors, as opposed to pointy incisors).  Many strepsirrhines (and tarsiers, too) have pointy incisors.  Williams et al. contended that the primitive state was to be more blunted and spoon-shaped, and that the pointy incisors were the ones that were derived.  Gingerich et al. counter that they “cannot imagine” that, although they can conceive of the fact that spatulate incisors are convergently derived in the taxa which display them (Darwinius, eosimiids, and some omomyids).
  • Absence of claws or grooming claws.  Lemurs, lorises, and tarsiers all have a grooming claw on their second digit.  Monkeys and apes do not, except for the callitrichines.  Tarsiers have a claw on the third digit, as well.  As for fossil adapiforms or omomyiforms?  Nobody knows. Fossil lemurs have been found without grooming claws, but none of have been found with a conclusive claw. Again, Gingerich et al. “cannot imagine” that nails on all digits could be primitive for primates.

What the hell?  Since when has “I can’t imagine!” been a valid scientific argument?  Especially since Williams et al. traced their line of reasoning so clearly in their paper!  Compare this exerpt from Williams et al.:

However, the polarity of this feature is unclear. Extant lemurs, lorises, and Tarsius have a claw on the distal phalanx of digit II of the foot, whereas non-callitrichine anthropoids do not (Soligo and Muller, 1999). Tarsiers also have an additional claw on digit III of the foot. Gunnell and Rose (2002) note that there is no clear evidence for the presence of a digit II claw in adapiforms and omomyiforms. A toilet claw has been reported for Europolemur kelleri ([van Koenigswald, 1979] and [Franzen, 1994]), although from our own observations of high-resolution casts of the specimen in question (HLD-Me7430) we remain unsure. Its congenerEuropolemur koenigswaldi lacks a toilet claw (Franzen et al., 2009). Thus, it is plausible that the presence of flattened nails on all digits is the primitive condition for euprimates and that toilet claws evolved independently in extant strepsirrhines, tarsiers, and callitrichines. Accordingly, the presence of a nail on digit II supports neither haplorhine nor anthropoid status for Darwinius.

with this exerpt from Gingerich et al.:

Claws or grooming claws on the pes change from present to absent between nodes 13 and 12 in the phylogram of Figure 2. Thus the loss of grooming claws and acquisition of nails on all digits in Darwinius is appropriately counted as a derived characteristic shared with Haplorhini.

Asserting something does not make it so!  Yes, based on the taxa that they used to construct their tree (mainly living taxa), there is a split between tarsiers and the rest of the haplorrhines.  But did they use the right traits?  Or the right taxa?  Make intelligent choices.  Don’t be bound by your “imagination.”  You can plug a bunch of living taxa into a computer program and then figure out where your fossil fits, but is that really making a sound scientific choice?  Don’t those other fossils count for something?

Figure 2 from Gingerich et al. 2010

Two more traits which Gingerich et al. consider to be derived are inclued in their article, but they are discussed as simply occurring between nodes and “thus” “appropriately counted” as derived features.

Gingerich et al. say that the Williams crew did not find any traits which were derived toward Strepsirrhines, but I think that may miss the point that Seiffert (2009) made with Afradapis- adapiforms are a sister clade to modern Strepsirrhines, and are derived in their own direction.

Am I missing something?  Why was this article so disappointing?

Gingerich, P., Franzen, J., Habersetzer, J., Hurum, J., & Smith, B. (2010). Darwinius masillae is a Haplorhine — Reply to Williams et al. (2010) Journal of Human Evolution DOI: 10.1016/j.jhevol.2010.07.013
Williams, B., Kay, R., Christopher Kirk, E., & Ross, C. (2010). Darwinius masillae is a strepsirrhine—a reply to Franzen et al. (2009) Journal of Human Evolution DOI: 10.1016/j.jhevol.2010.01.003
Seiffert, E., Perry, J., Simons, E., & Boyer, D. (2009). Convergent evolution of anthropoid-like adaptations in Eocene adapiform primates Nature, 461 (7267), 1118-1121 DOI: 10.1038/nature08429

About these ads

7 responses to “Darwinius massillae, continued…

  1. Eric September 18, 2010 at 1:03 pm

    I found this paper to be quite interesting, but mostly because it perfectly shows the problems you may encounter when building phylogenetic hypotheses.

    My first thought while reading was: “Why didn’t they use any other fossils species in their analyses?”

    Ok, without using any other adapids in their analyses, Darwinius becomes a sister group to extant anthropoids. But you can’t just exclude all the other fossils because they would mess up your tree.
    If their proposal is true and if Darwinius is still an Adapid, then you also have explain, why the other Adpapoidea are stem haplorrhines and why they lack certain traits, which are normaly related to this taxa.
    Their analyses doesn’t help to answer any of the general questions regarding the relationships of eocene primates. They just showed, that if they only use Darwinius and no other fossil primate from that period, it becomes a stem haplorrhine. And to be honest, I don’t think that the scientific value of such a proposal is very high.

    By the way, thank you for your post. It gave me something to work and think about for at least the whole weekend. Which is nice after serveral weeks of boredom.

  2. occamseraser September 23, 2010 at 2:21 am

    Good point, Eric. At the very least, they need to (and could) code an “Adapoidea” OTU, maybe an omomyid one too. But they also need to add another strep (like a mouse lemur) and another true anthropoid (perhaps a fossil OTU from the Fayum). Paup it again and see what happens.

    But I would have thought this is old turf for cladistics. I’m pretty sure that taxon-sampling and number of characters (and their interactions) have been explored in theory and practice. Any recent refs out there (none cited in Gingerich et al. — which BTW was still much better argued than I anticipated)?

    • Eric October 6, 2010 at 11:22 am

      I spent some time searching for References regarding the issue of taxon sampling in cladistic analyses (I was pretty bored recently). I only found two papers, one from 1998 and one from 2002. The first one is from David Hillis (Syst. biol. 47 (1) 3-8), the second one is from Derrick Zwickl and David Hillis (Syst. Biol. 51(4) 588-598).
      Both used molecular characters for their studies, so I don’t know how much we can apply their results to trees build with morphological characters. Anyways, both come to the conclusion that, espically when you deal with long branches, which you will do, when you mix fossil and recent species in your analyses, it’s better to increase the number of Taxa.
      The first paper also adresses the issue of investigator bias (Or “Theriot effect” asthe author called it) which I found quite interesting, especially if we look at the tree presented by Gingerich et al.

  3. Pingback: Afarensis: Anthropology, Evolution, and Science

  4. review upon September 19, 2011 at 4:29 am

    I read some part of the the pdf Human Evolution and the Media. I’m trying to correlate things with the above post. Need to complete reading the PDF before I can come out with my views.

  5. Pingback: Around the Web | TEMP Minds


Get every new post delivered to your Inbox.

Join 29 other followers

%d bloggers like this: