Re: giant birds

[Kendall Clements <k.clements@auckland.ac.nz>]

>Read more deeply into the systematics and the contradiction vanishes. 
>Don't forget that the birdlike theropods of the Cretaceous are many 
> millions of years removed from their Late Triassic through Late 
> Jurassic common ancestors with birds, and that they themselves have 
> undergone considerable evolution and adaptation into a lifestyle as 
> terrestrial bipedal cursors. Lots of convergence, too.

Character polarity is determined by outgroup comparison (at least in 
fossils), so a chronological gap in itself should not exclude a 
hypothesis. The detection of homoplasy (i.e. convergence) is presumably 
as much of a problem for BCF.
 
> I reject the "ground-up" hypothesis for the origin of avian flight on 
> physical grounds, because it is much, much less likely (astronomically less 
> likely) than a progressive "trees-down" origin. In the "ground-up" 
> hypothesis, ancestral birds do not acquire an arboreal lifestyle until 
> >after< they have evolve flight (otherwise "ground-up" and BCF differ only as 
> to the timing of the arboreality), and they evolve flight as cursorial 
> animals through the random accumulation of features that have nothing to do 
> with flying but, in some miraculous fashion, at some point in time become 
> exapted for flight. In BCF, birds go up into the trees, develop flight in at 
> least one lineage (and likely more) as a series of incremental improvements 
> to a progressively more efficient arboreal lifestyle. At various times--as 
> has happened repeatedly throughout avian evolution through the modern 
> era--some lineages abandoned the arboreal lifestyle and, if they had already 
> developed some kind of flying ability, became secondarily flightless, 
> cursorial animals. In the Mesozoic, we call such animals theropods; in the 
> Cenozoic, we call them phorusrachids, diatrymids, ostriches, and so forth.
> 
> We can use the theropod cladogram to track some of the evolutionary 
> developments of dinobirds as they became birds: bipedal posture was acquired, 
> the tail stiffened, the fifth and then fourth manual digits were lost to make 
> a rudimentary wing, the hallux retroverted, and so forth. BCF provides a 
> functional scenario for the evolution of bird flight that follows directly 
> from theropod phylogeny as it is currently understood.

The above scenario is reminiscent of an incremental, functional theory 
advanced by another George, George Gaylord Simpson. Ironically, unlike 
BCF Simpson's horse phylogeny was incompatible with fossil  
stratigraphy. Plausibility is a useful yardstick, but it depends on the 
eye of the beholder. A reading of the literature would suggest that not 
all find non-arboreal hypotheses implausible.

> "Proved wrong"? There's only one way to "prove" something in evolutionary 
> theory, and that is with a time machine. Tomorrow a newer hypothesis may 
> "prove" the "disproof" wrong. Cladistics tells us nothing about the nature of 
> the common ancestral forms, except that some of their features are apomorphic 
> for their descendant groups. It tells us nothing about which features may 
> have been present in the ancestral forms but were lost for functional reasons 
> in some of the descendant forms, when the only thing we have in the fossil 
> record are a few specimens of those descendant forms. BCF rejects the 
> obviously wrong parts of the "ground-up" and "trees-down" hypotheses and 
> synthesizes the remainder into a unified, parsimonious scenario for the 
> origin of avian flight. That's the best we can do for now, given the sporadic 
> (that is, nil) fossil record of Jurassic pre-Archaeopteryx dinobirds (for 
> which, by the way, BCF also provides a tentative explanation).

I'd say the finding that Acanthostega had (a) horizontally-directed 
forelimbs, (b) a non-load bearing pelvis that lacked strong attachment 
to the spine, (c) a true caudal fin with fin rays, and (d) functional 
internal gills pretty much excludes the possibility that it could get 
far on land. If nothing can be proved wrong in evolutionary theory, how 
can BCF conclude which are "obviously wrong" parts of theropod 
evolution and which are not? I'd much rather stick to an objective, 
repeatable methodology that does not rely on ad hoc reasoning (although 
I concede that character designation can be somewhat subjective). The 
critical question is one of methodology. BCF adherents prefer an 
incremental, functionally plausible narrative, while adherents of 
phylogenetic systematics prefer a repeatable, testable methodology 
based on character distribution. The latter requires that functional 
narratives conform to, rather than dictate, tree topology.

Now I must get back to marking undergraduate exams!

Kendall Clements

----------------------
Kendall Clements
k.clements@auckland.ac.nz