[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Genes show Neoaves branching before K/Pg extinction

David Marjanović <david.marjanovic@gmx.at> wrote:

> Hard or not hard, upper bounds _should_ be included much more often than
> they are. I've published on this (Laurin & me 2006, Syst. Biol.).

Oh. Many thanks for the ref!

> But the paper we're discussing didn't do that; it used minima without any 
> maxima.

It actually did; the supplementary material is clear on this. Out of
the seven calibration points, three had soft upper and lower bounds
(using Gaussian distributions with 95% of the overall probability
density located between the bounds), another three had hard lower
bounds with no upper bounds at all (lognormal distributions with
minima at the age of the fossils; their breadth was determined by the
default parameter values), and one apparently had hard upper and lower
bounds (the shape of the distribution used for this one isn't
mentioned in the text).

> You imply that Ratitae is defined by character states...

Of course -- it's even implied in the name itself. Is there any other
option? Gauthier & de Queiroz (2001) provided a phylogenetic
definition for the name, but I'm fairly sure that hardly any
ornithologists have ever heard of it. It's also impractical, as it
would refer only to _Struthio_ (and maybe _Palaeotis_) under the
current phylogenetic hypothesis.

Tim Williams <tijawi@gmail.com> wrote:

> The diagnosis of this ratite+tinamou clade (= crown palaeognathans)
> depends on the topology (unless it has an apomorphy-based definition).
>  If the ratite morphology originated more than once, then this
> morphology cannot diagnose the clade.

Yes, it cannot diagnose ratites as a clade -- but everyone agrees that
ratites are not a clade. I don't understand how the diagnosis of crown
paleognaths is relevant to this.

> I would have thought you could indeed distinguish paraphyly from
> polyphyly based on topology alone.

+-- Lepidosauria
`--+-- Aves
   `-- Crocodylia

Reptilia includes Lepidosauria and Crocodylia.

+-- Mammalia
`--+-- Lepidosauria
   `-- Aves

Haematothermia includes Mammalia and Aves.

The members of Reptilia and Haematothermia have exactly the same
position relative to the non-members on these trees, yet the first
group is paraphyletic (because the last common ancestor of lepidosaurs
and crocodiles was an egg-laying ectotherm) and the second one is
polyphyletic (because the last common ancestor of birds and mammals
was not an endotherm). I can't interpret it in any other way than that
the topology of a tree isn't enough to distinguish the two.

> The ancestral state is inferred
> from the topology.  The topology is itself based on the distribution
> of character states among the included taxa.

It's not just the topology -- ancestral state reconstruction methods
can use some other data, too, such as branch lengths (and a prior
probability distribution must be specified in Bayesian inference). A
method that assumes that changes are equally probable in both
directions can infer a different ancestral state than a method that
does not, even if they are both applied to the same topology.
Sometimes, our strong prior beliefs about the probability of a change
from one state to another trump the topological information, which is
probably why you admit that multiple losses of flight are more likely
than a single regain in tinamous (although the distribution of
character states in the terminal taxa would favor the latter

> According to one study (Alvarenga, 2010) _Diogenornis_ is closer to
> casuariids than to rheids.

... and Alvarenga should know, as he described it as a rhea in the first place.


Gauthier JA, de Queiroz K 2001 Feathered dinosaurs, flying dinosaurs,
crown dinosaurs, and the name "Aves". 7-41 _in_ Gauthier JA, Gall LF,
eds. _New Perspectives on the Origin and Early Evolution of Birds:
Proceedings of the International Symposium in Honor of John H.
Ostrom._ New Haven: Peabody Mus Nat Hist, Yale Univ

David Černý