. Bulletin of the Museum of Comparative Zoology at Harvard College. Zoology. Natural History Cerion VIII: A Revision • Gould and Woodruff 383. Figure 6. Position of mean vectors for all samples of Little Bahama Bank cerions. This is a triaxial plot of nor- malized factor loadings for a 3-axis, varimax solution in the Q-mode. These three axes explain per cent of all information; the first two axes explain 88 per cent. Ribby and mottled morphotypes are well separated by the first two axes. Mottled samples from Grand Bahama Island have higher projections on the third axis. Closed circles are
. Bulletin of the Museum of Comparative Zoology at Harvard College. Zoology. Natural History Cerion VIII: A Revision • Gould and Woodruff 383. Figure 6. Position of mean vectors for all samples of Little Bahama Bank cerions. This is a triaxial plot of nor- malized factor loadings for a 3-axis, varimax solution in the Q-mode. These three axes explain per cent of all information; the first two axes explain 88 per cent. Ribby and mottled morphotypes are well separated by the first two axes. Mottled samples from Grand Bahama Island have higher projections on the third axis. Closed circles are mottled samples from Abaco; open circles are mottled samples from Grand Bahama; crosses are ribby samples; squares represent samples defined by geography and ecology (not morphology) as inhabitants of zones of interaction between ribby and mottled populations (note their intermediate position in morphology as well); the star represents the single fossil sample from Abaco. The line connects samples of the hybrid zone at Rocky Point in geographical order, pc is the Pongo Carpet sample (mottled, partly convergent upon ribby); f is the fossil sample; cry are "C. chrysaloides" (the name applied to ribby samples on Grand Bahama); luc is "C. lucayanorum" (ribby sample from Mores Island). Other numbers refer to localities discussed in text. because it can control so much covariance in a matrix (large shells have high values of almost all variables), and because all mor- photypes and areas contain both large and small-shelled samples. We eliminate this pervasive control of size in order to see smaller but more stable influences more clearly. However, we also performed an analysis without normalization and obtained nearly identical results (see below). Three axes encompass per cent of the information in 52 samples; no subse- quent axis reaches one per cent. We per- formed a varimax rotation and computed the factor loadings of all samples upon the three axes (i
Size: 1696px × 1473px
Photo credit: © Book Worm / Alamy / Afripics
License: Licensed
Model Released: No
Keywords: ., bookauthorharvarduniversity, bookcentury1900, booksubjectzoology
