-------- Original Message -------- Subject: Using regression residuals to remove geomorphological bias Date: Thu, 17 Apr 2008 09:11:19 -0700 (PDT) From: Craig Farquhar <[EMAIL PROTECTED]> To: <[email protected]> I have a dataset of morphometric variables measured from museum specimens which were collected over a wide latitudinal and altitudinal range (across north-south range of the South American Andes). I'm interested in looking at potential boundaries in morphometric space that could diagnose species level differences for taxonomic purposes. Because of the very real possibility that size differences could be due solely to geomorphological bias (e.g., Bergmann's Rule) I sought to eliminate this by comparing residuals from linear regressions of a given morphometric variable (e.g., lengths of primaries, tail, tarsus, etc.) against altitude and latitude. But I have been criticized in doing this because it would cloud the very species level differences I might be looking for. My opinion is that true species level differences would be found irrespective of geomorphological bias. In fact, this statistical approach was suggested to me upon consultation with the late Dr. Leslie Marcus whose opinion I've always greatly respected. Has anyone out there had similar experiences trying to decide whether to remove geomorphological bias from a taxonomic study? Is it really a bias? I'd love to hear any and all constructive comments along these lines. Craig ____________________________ Dr. C. Craig Farquhar Section 6 Grant Program Coordinator Wildlife Division Texas Parks and Wildlife Department 4200 Smith School Road Austin, Texas 78744 Office: (512) 389-4933 Fax: (512) 389-8043 Email: [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> -- Replies will be sent to the list. For more information visit http://www.morphometrics.org
