This paper exhibit average values, standard error and Pearson correlations of 10 linear measurements (withers, rump and thorax height; shin and thoracic perimeters; rump and body lengths; opening of ischium; breadth of the rump, and body weight) of 14 heifers Pantaneira breed, as well as a multivariate analysis using principal component analysis (PCA) to reduce the amount of variables that can be kept and used in selection, due to a higher contribution to the total variation. The animals, with 36 months, not pregnant, belong to Nucleus of Cattle Pantaneiros of Aquidauana, University State of Mato Grosso do Sul. The heifers presented measures near the small sized cattle, such as Sindhi and Jersey breeds, with weight of 375±9.95kg. The body weight showed a moderate correlation and significant (P<0.05) with withers (55%) and rump heights (69%) and body length (59%), and high thoracic perimeter (82%). Therefore, by choosing the highest exemplary, of longer and greater rump and thoracic circumference, certainly will be selected the heavier animals. Based on the weight of the measures were needed only 4 PCA with minimum autovalue of 1.0, reaching 83.95% of the cumulative variance. It was extracted 1 characteristic with the highest weight in each PCA, which may be maintained in future studies or screening programs, with a reduction of 60%, thus, indicating the measurements of height and breadth of the rump, thoracic perimeters and length body as being essential to select of animals phenotypically superior.
Bos taurus taurus; conservation; measurement; naturalized breed; selection