Farewell to Species - reticulation

[Pierre Deleporte]

Comment to Les Kaufman (text below):

To my sense, constructing reticulate networks is not obligatorily at odds
with cladistic thinking, as long as hennigian phylogenetic systematics (and
thus some way... evolutionary systematics) is concerned. The problem is to
identify deviations from the basic evolutionary model of descent of
characters in direct lineage, which supports phylogeny reconstruction methods.
 
John ALROY proposed CTA (Continuous Track Analysis software), which starts
with a classic cladogram (optimizing character states contiguity), and
allows to identify the instances of reticulation saving the more
homoplasies. If a few reticulations explain a lot of homoplasy, these are
the best candidates for instances of past hybridization... under the model
of hybridization = massive exchange of characters between the two lineages.
Of course, as Alroy correctly stated,  an abuse of such explanations will
trivially resolve all character conflicts by ad hoc hypotheses of
hybridization.

The problem is: does hybridization generally leaves such landmarks behind?
More generally: what are the specific traces, if any, left by the process
of hybridization in lineages? For instance, is homoplasy due to
hybridization differentiable from homoplasy due to other processes? (by its
distribution? by the quality of characters involved?...). Same question for
lack of cladistic resolution (quick radiation of generalized
hybridization?). What experimental or populational field data can support
these guesses?

A question: what makes you suspect reticulation in your Cichlid fishes?
Your answer to this question should hopefully contain the keys for
identifying hybridization-reticulation.


Pierre Deleporte (a guy)


A 08:03 01/02/00 -0500, you wrote :
>Hi, guys, need some help here.
>
>Cichlids again.  Field data and genetics strongly suggest that we're
>looking at reticulate evolution.  Schluter et al.'s recent paper in
>Science on parallel speciation in sticklebacks provided experimental data
>to support the plausibility of my contentions, published earlier.  Time to
>take network phylogenies seriously.
>
>Has anybody out there thought a lot about this, and have recommendations
>for the best available protocols for non-Hennigian...can I even use the
>term...cladistics?  Have the bad-boy aspects of Mayr's evolutionary
>systematics ever been properly formalized?
>
>Les Kaufman
>Boston University Marine Program
>Department of Biology
>5 Cummington Street
>Boston, MA 02215
>lesk at bio.bu.edu
>617-353-5560 office
>617-353-6965 lab
>617-353-6340 fax
>
>
Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 66
Télécopie : 02 99 61 81 88