Biostatistics Research Today is a free monthly online journal that collates and summarizes the latest research about Biostatistics, including details on statistics, uncertainty, probability, modeling.

Is geometric morphometrics efficient for comparing otolith shape of different fish species?

Ponton D

IRD, BPA5, 98848 Nouméa, New Caledonia. Dominique.Ponton@noumea.ird.nc

The 2D shape of sagittae of Encrasicholina devisi, E. heteroloba, E. cf. punctifer, and Stolephorus indicus, four tropical Engraulididae of New Caledonia, was studied with 1) dimensionless shape descriptors (form factor, roundness, and aspect ratio); 2) elliptic Fourier analysis (EFA) and Fast Fourier Transform (FFT) of 2D outline; and 3) geometric morphometrics (GM) based on four standard landmarks and nine semi-landmarks. The largest sagittae of E. cf. punctifer were lacier, i.e., had a smaller form factor, than the sagittae of the other species. The sagittae of E. devisi and S. indicus were more roundish, and presented a lower aspect ratio, than those of E. heteroloba and E. cf. punctifer. Between-class correspondence analysis (COA) indicated that between-species inertia was the lowest when based on the 96 Fourier coefficients originating from EFA, and the highest when based on the 22 partial warps originating from GM. As otoliths of different sizes from different species presented similar shapes, relative between-species inertia increased markedly when length, width, perimeter, and area were added to the set of variables originating from EFA, FFT, or GM. Despite otoliths having only a few, sparsely located, homologous landmarks, GM appeared slightly more efficient in distinguishing the sagittae of the four species and allowed visualization of the modification of otolith shape as they grow.

Published 3 May 2006 in J Morphol, 267(6): 750-7.
Full-text of this article is available online (may require subscription).

© 2005-2008 Biostatistics Research Today. All Rights Reserved.