-------- Original Message -------- Subject: e-mail for the list Date: Sat, 23 Jan 2010 15:10:12 +0100 From: [email protected] To: morphmet <[email protected]> Dear morphmet community, I have no experience in morphometrics so far. I want to use morphometric methods in my diploma thesis to answer some questions about two mussel species. My list of literature gets bigger and bigger, but I cannot find all answers and cannot understand some steps in data processing. That´s why I ask for help here. I already made pictures of >250 mussels (2 species, 2 and 4 localities) and used tpsdig2 to create the outlines. It is impossible to use landmarks because of the lack of homologous points/structures. That is the reason, why I can ?only? use EFA to answer my questions. But before I create a larger dataset, I have to understand which methods to use, to answer my questions. The first question (creating data from outlines with PAST): 1. Can I use ?procrustes (2D) + rotate to major axis? for outlines (1st step)? (In my literature procrustes is usually mentioned in connection with landmarks.) Aim of my study: I want to find out, if there is more than one morphotype within each of these two species. If there are more morphotypes, I want to test if this depends on different variables like physico-chemical conditions (and so on). Furthermore, I want to test, if these morphotypes are present in all sampled locations or if the distribution of morphotypes is limited to some of the sampled locations. The study of Marquez et al. (2010) contains several methods of interest for my study. I want to use approximately the same methods. Marquez et al. (2010) used multivariate regression (as suggested by Klingenberg (2008)) to detect and remove the effect of size (or other environmental factors) before shape comparison between groups of potential different pheno-( or morpho-)types. The cited link to Klingenbergs´ homepage is not available and I cannot find out, what exactly Klingenberg suggested. Some more questions: 2. Can I make such a multivariate regression in PAST with more than 1 variable (like length, width, height, Calcium-content, water temperature and so on)? How can I test allometric growth in this context? 3. How can I remove these effects from the outlines? 4. Should I first use linear regression to test allometric growth (length, width, height)? Can I make cluster analysis to find groups of individuals with a similar isometric or allometric growth? Does this make sense at all? I hope that you can help me to learn more about geometric morphometrics. Please answer me directly to my e-mail address or if there is literature I should read to answer my questions, send me links. I am grateful for each support! [email protected] Best regards, Sascha Brandt ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. -- For now, new message AND replies should be sent to: [email protected] /* Replies will be sent to the list. */ For more information visit http://www.morphometrics.org
