出版社：Universidad Nacional del Centro de la Provincia de Buenos Aires
摘要：El objetivo de este trabajo es comparar técnicas derivadas de la morfometría tradicional y geométrica para el análisis de la forma del esqueleto facial a nivel intraespecífico. Como caso de estudio, se analizaron tres muestras diacrónicas de individuos masculinos del Sudeste de la Región Pampeana y dos del Nordeste de Patagonia (Barrientos y Perez 2002). Sobre las variables obtenidas a partir de la aplicación de las distintas técnicas (proporciones, residuos multivariados, residuos Procrustes y deformaciones parciales), se realizaron análisis de componentes principales, correlación producto- momento de Pearson, ANOVA de clasificación única y test post-hoc de la varianza (Least Significant Difference o LSD). Se obtuvieron tres conjuntos de resultados representados por: a) los residuos multivariados, b) las proporciones y los residuos Procrustes, y c) las deformaciones parciales. Las diferencias observadas entre tales conjuntos son consistentes con los resultados obtenidos en recientes evaluaciones parciales e independientes de las técnicas utilizadas en este trabajo. Los mismos indican que los residuos multivariados son poco eficientes para separar el componente tamaño del componente forma (Jungers et al. 1995), que las proporciones y los residuos Procrustes describen el cambio global de la forma (Jungers et al. 1995; Lynch et al. 1996), y que las deformaciones parciales proporcionan mayor cantidad de información morfométrica con significado taxonómico ya que describen las diferencias en la posición relativa de cada uno de los landmarks considerados en el análisis (Rohlf et al. 1996).
其他摘要：The aim of this paper is to compare different techniques derived from traditional and geometric morphometrics used to analyze the shape of the cranio-facial skeleton at an intra-specific level. Three diachronic samples belonging to male individuals from the southeastern Pampas were taken as the case study. They belong to the Early/Middle Holocene transition (Ht/m, ca. 8000 to 6000 B.P., n= 4), the Early Late Holocene (Htai, ca. 3000 to 2000 B.P., n= 6), and the Later Late Holocene (Htaf, ca. 1000 to 400 B.P., n= 4) (Barrientos and Perez 2002). The other two samples were taken from the northeast of Patagonia, and although there is no radiometric chronological data, they probably belong to the Late Holocene [e.g. Isla Gama (IG), n= 6 and San Blas (SB), n= 8] (Figure 1 and Table 1). These samples were chosen on the basis that they are currently being used to study the peopling of the southeastern Pampas (Barrientos and Perez 2002). Principal components analyses were performed on the variables obtained by using several statistical techniques (ratios, multivariated residuals, Procrustes residuals and partial warps), Pearson product-moment correlation, one way ANOVA and a variance post hoc test (Least Significant Difference or LSD). Three sets of results were obtained, and they are represented by a) the multivariate residuals, b) the ratios and Procrustes residuals and, finally, c) the partial warps. The differences registered between these sets are consistent with the results obtained in a recent partial evaluation made of the techniques employed in this paper. They indicate that the multivariate residuals are not so efficient for separating the size component from the shape component (Jungers et al. 1995), that the ratios and Procrustes residuals describe the global shape change (Jungers et al. 1995; Lynch et al. 1995), and that the partial warps offer a greater amount of taxonomically meaningful morphometric information, since they describe the differences in the relative position of each landmark considered for the analysis (Rohlf et al. 1996).
