Heather,
Sexual dimorphism is the difference between males and females, so one
can quantify this as the difference between average phenotypes for each
sex. This literally is a vector in shape space connecting the mean male
and mean female shapes. The length of this vector describes how much
sexual dimorphism is present, and its direction in shape space describes
how shape differs between males and females.
If one has multiple populations, a 2-factor MANOVA with population, sex,
and pop X sex interaction allows a quick assessment of whether or not
populations differ in their sexual dimorphism. This would be found by a
significant interaction term. If the interaction term is significant,
then it may be that the magnitudes of their SShD vectors, or their
directions (or both), differ among populations. To determine this, one
must directly examine the vector magnitudes and directions statistically.
A few years ago we proposed a procedure for comparing these multivariate
phenotypic change vectors, which we subsequently generalized to
phenotypic trajectories. The relevant papers are: Collyer and Adams 2007
(Ecology); Adams and Collyer 2007 (Evolution); Adams and Collyer 2009
(Evolution). The papers and R-code for implementing the procedure can
be found on my web page. Also, one of the examples in the 2007 Ecology
paper compares sexual dimorphism vectors across several populations of
pupfish. Another sexual dimorphism example is found in Berns and Adams
2010 (Auk).
Hope this helps.
Dean
--
Dr. Dean C. Adams
Associate Professor
Department of Ecology, Evolution, and Organismal Biology
Department of Statistics
Iowa State University
Ames, Iowa
50011
http://www.public.iastate.edu/~dcadams/
On 7/26/2010 11:50 AM, morphmet wrote:
-------- Original Message --------
Subject: morphometric measures of dimorphism
Date: Sun, 25 Jul 2010 12:44:39 -0400
From: [email protected]
To: [email protected]
Hi All,
I have a project in which I've extracted the browridge area of crania
and I'm interested in male and female differences in shape, and how
these differences vary across populations. Ideally, I'd love to use
sliding landmarks on the 3D surface, but I can't seem to come across a
program to help me out with that. In the mean time I've taken some
transects and have equally spaced points along these curves to describe
the overall shape. I've done the Procrustes superimposition and the PCA
and it looks to pick up the shape differences I'm interested in and
separates males and females well.
What I'd really like to do, however, is have some kind of measure of
dimorphism that I can compare to other populations, to determine whether
some populations are more dimorphic in brow shape than others. For
example in stature the most common method is taking the ratio of average
male stature to female. Any ideas of the best way of doing this? I
thought maybe if I conducted a discriminant function for all populations
pooled, then maybe I could compare the average score of males and
females from each sample???
Thanks in advance for any insight and guidance!
Heather Garvin
[email protected]
