-------- Original Message --------
Subject: for morphmet: advice on using eigenshape scores to calculate Qst?
Date: Tue, 27 Jul 2010 21:47:42 +0100
From: Kisel, Yael <[email protected]>
To: [email protected]
<[email protected]>
Hi all,
I'm a PhD student trying to finish up my thesis on population genetics &
diversification. One of the last analyses I want to do is to use
morphometric data from leaf outlines to calculate Qst's/Pst's for my
study species, which are all tropical orchids.
My plan at the moment is to use either standard eigenshape or extended
eigenshape analyses to get independent axes of shape variation, then
calculate the within- and between- population variation of these axes
for each species, and from that calculate Qst (or Pst, since I have no
information on heritability).
I don't have much background in morphometrics, so I would really
appreciate some advice:
- All of my study species have quite simple, elongate leaves. I am
interested only in general shape, rather than asymmetry or anything else
like that. So I was thinking to use just half outlines - the outside
curve of one side of each leaf, from base to tip. Is this reasonable, or
would I gain anything by using the outline of the whole leaf?
- What is the best way to calculate Qst from morphometric data? Every
paper I read seems to use a different equation or a slightly different
method. What are they key references I should read?
- Is there any program/R package that does Qst calculations?
- How do I deal with the multiple shape axes/principal components axes
that I will get for each dataset - do I need to calculate Qst separately
for each?
- Is there anything else I should keep in mind or reconsider?
You can reply to me directly at y.kisel06 (at) imperial.ac.uk
Thanks very much!!
Yael
--------------------------------------------
Ms. Yael Kisel
PhD student
Royal Botanic Gardens, Kew &
Imperial College, London
Silwood Park Campus
Ascot, Berkshire SL5 7PY UK
