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

Re: New Phylogeny of Birds in PLoS ONE

Ronald Orenstein <ron.orenstein@rogers.com> wrote:

> Tis is probably because I don't really understand the nitty-gritty of these 
> sorts of analyses, but why only 32 species?

A higher number of species would probably make the analysis
computationally unfeasible. With data sets this large, the
computational power required to complete even a regular ML or Bayesian
analysis might be prohibitively high -- Hackett et al. (2008), whose
data set was about three times smaller than the current 1541 locus
one, needed months of supercomputer time for their analyses and it
still wasn't enough for their Markov chains to reach convergence.
Recent methods for inferring species trees from discordant gene trees,
such as the one used by McCormack et al. (2013), make the situation
even worse. The authors explain their use of a relatively small number
of taxa in the discussion:

"Our sampling strategy sought to balance the number of taxa included
with the number of loci interrogated. We sampled the genome much more
broadly than the 19 loci of Hackett et al. [6], but with reduced
taxonomic sampling (32 species compared to 169 species)."

> No Anseriformes, for example! (OK, the paper is about Neoaves, but still...) 
> No rails, sandpipers, or gulls; no herons, ibises or storks (if you don't 
> count the hamerkop and shoebill); no turnicids; etc.

I don't think the omission of these taxa is a big problem, as
monophyly of Neoaves, core gruiforms, and charadriiforms is pretty
well established. Cuckoos, seriemas, mesites, and the cuckoo roller,
on the other hand...

evelyn sobielski <koreke77@yahoo.de> wrote:

> Essentially, what I found is "landbirds" and to a lesser extent Aeqornithes 
> internal phylogeny switches between 3 more-or-less discrete states depending 
> not on character sample but on taxon sample and differing most significantly 
> in the position of Passeriformes.

Isn't the support for their sister group relationship to parrots
rather robust, though? It's even one of the few results that emerged
from McCormack et al.'s species tree analysis, despite its overall
lack of resolution.


Hackett SJ, Kimball RT, Reddy S, Bowie RC, Braun EL, Braun MJ,
Chojnowski JL, Cox WA, Han K, Harshman J, Huddleston CJ, Marks BD,
Miglia KJ, Moore WS, Sheldon FH, Steadman DW, Witt CC, Yuri T 2008 A
phylogenomic study of birds reveals their evolutionary history.
Science 320(5884): 1763-8

David Černý