Using form analysis techniques to improve photogrammetric mass-estimation methods

摘要

Numerical characterization of animal body forms using elliptical Fourier decomposition may be a useful analytic technique in a variety of marine mammal investigations. Using data collected from the Weddell seal (Leptonychotes weddellii), we describe the method of body form characterization using elliptical Fourier analysis and demonstrated usefulness of the technique in photogrammetric mass-estimation modeling. We compared photogrammetric mass-estimation models developed from (1) standard morphometric measurement covariates, (2) elliptical Fourier coefficient covariates, and (3) a combination of morphometric and Fourier coefficient covariates and found that mass-estimation models employing a combination of morphometric measurements and Fourier coefficients outperformed models containing only one covariate type. Inclusion of Fourier coefficients in photogrammetric mass-estimation models employing standard morphometric measurements reduced the width of the prediction interval by 24.4%. Increased precision of photogrammetric mass-estimation models employing Fourier coefficients as model covariates may expand the range of ecological questions that can be addressed with estimated mass measurements.