Predicting the evolution of Theropod arm size

In 1999 I published an article in the (cryptic) monthly news bulletin of the Museum of Natural History of Chile, entitled "Evolution of Arm Size in Theropod Dinosaurs: A Developmental Hypothesis". In this work, which I did as an undergrad, I tackled what seemed a consistent trend among species of theropod dinosaurs, the presence of proportionally smaller arm sizes in species with larger body size. Femur size correlates quite directly with body size in dinosaurs: I therefore took femur size as an indicator of body size and making a linear regression I obtained that Humerus = 0,387 (Femur)^0,6807, with a rather nice determination coefficient, R^2 = 0,89. It is, of course, very remarkable that this negative allometry would so nicely predict theropod humerus size. Rather than the usual adaptive explanations, I presented a developmental hypothesis: The ontogeny of the most recent common ancestor of the theropods presented slower growth of the arm than of the leg, with arms that were proportionally larger earlier in ontogeny, at smaller body sizes. Thereafter, evolutionary variation of different body sizes in different species descended from this ancestor simply reflected conservation of this ontogenetic trait, with smaller- or larger-looking arms as a consequence of body-size variation, rather than any process of "local" adaptation. In this context, the very bird-like dromaeosaurid dinosaurs would owe their proportionally large arms to their small body size. If this were indeed correct, if a dromaeosaurid species would evolve a larger body size, we would expect it to show short arms, unlike those of its small-sized, long-armed ancestors.

The recent discovery of the dromaeosaurid Austroraptor is just that: a pretty large dromaeosaurid, with proportionally small arms, despite being descended from long-armed, smaller ancestors.

So, how does the Vargas formula work for Austroraptor? The femur length of Austroraptor is 0,56 M, so the predicted humerus length is 0,261 M. The actual length of Austroraptor's humerus is 0, 262 M... not bad, hehehe.

Of course, my hypothesis has another more direct and straight-forward prediction: Within an ontogenic sequence of a theropod species, younger, smaller specimens should present proportionally larger arms than older, larger specimens. Unfortunately, I still don't know of any conclusive evidence for this in a non-avian theropod. However, this is known to have been the case for Prosauropod dinosaurs, the inmediate outgroup of theropods, which presented larger arm porportions at earlier stages of ontogeny. Furthermore, according to Middleton and Gatesy, there is preliminary data that forelimb size decreases in the gull Larus californicus form 43% in hatchlings to 38% in adults...suggesting, by phylogenetic "sandwich", that the same was true for non-avian theropods. I thus renew the validity of my prediction!

Vargas A 1999 Evolution of Arm Size in Theropod Dinosaurs: A Developmental Hypothesis. Noticiario Mensual del Museo Nacional de Historia Natural de Chile 338: 16-19

Novas, F. E., Pol, D., Canale, J. I., Profiri, J. D., Calvo, Jorge O., 2008. A bizarre Cretaceous theropod dinosaur from Patagonia and the evolution of Gondwanan dromaeosaurids. Proceedings of the Royal Society B.

Evo-Devo: The Good, the Bad, and the Ugly.

THE GOOD: Homology assesment evo-devo, and epigenetic evo-devo. Evo-Devo that works on the pretty empirical task of assessing problematic homologies, with tree-based inferences on the evolution of development, continues to greatly help the reconstruction of the evolutionary history of life on earth. Epigenetic evo-devo's, in turn, understand that developmental biology is not genetics. They have realistically confronted the role of higher level and environmental epigenetic interactions in development, and thus also in the origin of evolutionary novelties. Both of these tend to emphasize how standing developmental mechanisms, and not natural selection alone, are essential to the pathway taken by evolution.

THE BAD: Reductionist "regulatory" evo-devo. Dangerous, because it is is upheld by important figures of evo-devo, presenting itself as a triumph and empirical conclusion. Yes, mutations in cis-regulatory regions are commonplace in evolution, but these people seem to have reductionist difficulties in understanding there is anything more beyond finding such a mutation. The notion that only non-coding "regulatory" sequence changes can produce localized expression (in time or space) simply makes no good developmental sense. As Lillie pointed out, all cells have the same DNA content, including the "regulatory" elements; whether a gene is expressed or not still varies from cell type to cell type depending on something else as well. In other words, Lillie's "paradox" forces the question: "who regulates the regulators"? This question reveals that "regulation" is nothing but a sloppy, semi-nonsensical wastepaper-basket term. Both coding and non-coding sequences can be "regulatory". Even environment can "regulate" gene expression. Genes are expressed differentially in cells, NEVER because of their "regulatory" sequences alone, but ALWAYS including the higher-level and environmental interactions at the cell and tissue level, which explain "Lillies paradox". (again: This is why developmental biology is different from genetics!). Focusing only on one type of mutations (cis-regulatory) is just a re-strengthened version of the old reductionist fallacy that genotype=phenotype. This false equivalence ultimately downplays the role of understanding development, the actual mechanisms that relate genotype to phenotype. Without really introducing developmental mechanisms, no serious challenge is made to the hegemony of population genetics as a way of understanding evolution. Yes: The bad is a traitor of development, for love of genetics. Evo-Devo can now become a mere footnote: The largely uninteresting filling-in of superfluous data on "what the specific mutations were".

THE UGLY. Just plain wrong or artifactual topics, mostly born from the lack of proper integration of different fields of research. Specially silly is the "conflict" between microevolution and macroevolution. Doubtless, the study of microevolution offers many advantages. But this does not mean at all that a macroevolutionary study will not be able to derive sound conclusions: when the evidence is there, there is nothing to say about the micro or macro level in which a question is satisfactorily answered. That comparable experiments can only be tested between closely related species is a myth: gene expression experiments can produce the same phenotypic alteration despite hundreds of millions of years of separation.
This follows in an old lab-bench tradition of being purely "experimental" negating any need to know much about natural history and macroevolution. It also relates to "blind" faith in molecular phylogenies, that is, with little capacity for critically evaluating these studies (such as by morphologica implications). As a result, plain artifacts of the tree become the basis for many weird hypotheses (I have argued before this is happening right now, with new supposed clades such as "urochordates+vertebrates"). Studies continue to emerge where well-established facts of natural history are swept aside in favor of some "groundbreaking hypothesis".

Many Evo-Devos have thought that macroevolution is some kind of truly radical "body plan" change with mechanisms quite different from those observed at a microevolutionary level, appealing to some mysterious happenstance of the past for the origin of phyla or higher "grades". However, Evo-devo's are slowly wisening up to the fact that macro and micro evolutionary change proceeds pretty much in the same way, with the simple fact being that lineages diverging earlier can accumulate greater differences (as pointed out before)