SEX DIFFERENTIATION OF THE BAIRD SPARROW (Centronyx bairdii) THROUGH ARTIFICIAL NEURAL NETWORKS AND MORPHOMETRIC DATA

Abstract

Estimates indicate that 148 species of birds in North America face a high or severe threat, because their populations are decreasing. The Baird sparrow (Centronyx bairdii) is one of the species with population decline, it presents monomorphic plumage and sex cannot be determined at plain sight. Sex determination in birds allows understanding their social behavior and proportion regarding population dynamics. Artificial neural networks are used as a classification method which has been used in various fields, such as plant and seed classification, and species differentiation of timber species. In the scope of ornithology, it has been scarcely used. Therefore, the objective of this study was to develop an artificial neural network to differentiate the sex of sparrows of the species Centronyx bairdii, with zoometric data obtained in the field. The hypothesis was that artificial neural networks can predict the sex of the birds of the species Centronyx bairdii. The neural network built from the data of weight, wing chord, tail length, culmen, width of the beak and depth of the beak of 111 birds, allowed differentiating the sex of the species with a degree of certainty of 92.3%.