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

Re: Bird diversification through time

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

>> Here we present, analyse and map the first complete dated phylogeny
>> of all 9,993 extant species of birds
> That's a supertree, right? :-/

No, it's a result of a series of partitioned phylogenetic analyses.
The authors assigned all 6,663 species for which genetic data were
available to 129 smaller clades whose monophyly was assumed
(constrained). Both the relationships between and within the clades
were inferred using BEAST, with backbone constraints based on
well-supported nodes from Ericson et al. (2006)* and Hackett et al.
(2008) for the higher-level phylogeny and no constraints within the
clades. The positions of taxa that weren't represented by sequence
data were inferred using taxonomy as a proxy for phylogeny: Jetz et
al. assumed that all genera were monophyletic unless there was some
molecular evidence for their non-monophyly, in which case they
constrained the taxa deeper in the tree or nor at all (genera without
any genetic data were treated the same way). They combined all those
data to assemble the final tree with 9,993 tips -- about 18% of the
nodes were constrained in some way, but the rest was estimated from
actual sequence data.

*They used the tree from the Supplementary Figure 6 (all genes except
beta-fibrinogen) to account for the potentially distorting effects of
FGB-int7, which is pretty cool.

> ...One does not simply calculate diversification ( = cladogenesis) rates
> from a tree that only contains extant taxa. That would amount to assuming an
> extinction rate of 0, which is (alas) empirically wrong.

There are purely neontological diversification studies that use models
with a non-zero extinction rate estimated as an additional parameter,
though (Rabosky & Lovette 2008). Of course, the pure birth (no
extinction) model can still be used if it fits the data well.

> Does the paper try to do anything about this problem?

It does; I don't think it would be publishable if it just assumed the
pure birth model and didn't compare it to more sophisticated models.
>From the supplementary info:

"We then evaluated the fit of nine models of diversification using an
AIC framework. Using LASER, we assessed pure birth, constant-rate
birth-death, density-dependent logistic (where extinction is set to
zero and speciation rate follows a logistic curve as a function of
standing diversity (Sepkowski 1978, Rabosky and Lovette 2008a)),
density dependent exponential (where extinction is set to zero the
speciation rate is a function of the number of extant lineages at any
point in time) and two more complex models (SPVAR and EXVAR; Rabosky
and Lovette 2008b)."


Ericson PGP, Anderson CL, Britton T, Elżanowski A, Johansson US,
Källersjö M, Ohlson JI, Parsons TJ, Zuccon D, Mayr G 2006
Diversification of Neoaves: Integration of molecular sequence data and
fossils. Biol Lett 2(4): 543-7

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

Rabosky DL, Lovette IJ 2008 Explosive evolutionary radiations:
decreasing speciation or increasing extinction through time? Evolution
62: 1866-75

David Černý