Sample size, hybrids, and R

[morphmet_moderator]

----- Forwarded message from "Samuel F. Rizza" <samuel.ri...@humboldt.edu> -----

Date: Tue, 11 Mar 2014 20:34:49 -0400
From: "Samuel F. Rizza" <samuel.ri...@humboldt.edu>
Reply-To: samuel.ri...@humboldt.edu
Subject: Sample size, hybrids, and R
To: morphmet@morphometrics.org

Hi all,

I am a graduate student looking at morphometrics and genetics of coastal cutthroat trout(CCT), steelhead trout(SH), and their hybrids(HY).  I used 14 landmarks on 381 CCT, 182 SH, and 15 HY (these are putative species IDs recorded in field until genetics are determined).  I am trying to determine if HY are morphologically different than CCT and/or SH and if so, how?  I understand this may be hard with such a small HY sample size.    

1. Does the difference in sample sizes of CCT, SH, and HY alter the consensus shape effecting my results when comparing group differences in PCA and CVA?

If so, 

2. Is R the best program to use to randomly sample an equal number of CCT from a TPS file?

3. How would I go about comparing differences between CCT, SH, and HY morphologies from the headwater reaches vs. lower river/estuary? My hypothesis being lower river/estuary fishes have more hybridization and different life histories resulting in less differentiation of consensus shapes of species.

I am most familiar with MorphoJ, IMP7/8, and tpsRelw.

 

Thanks for all the posts, love the forum.  Feel free to contact me directly with any papers,programs to try, or suggestions. Just getting into the field of morphometrics and enjoying its capabilities.  

Thanks,

Sam Rizza

Fisheries Graduate Student

Humboldt State University, CA

sf...@humboldt.edu

----- End forwarded message -----