Recomendação de acasalamentos para otimização de modelos bioeconômicos

Abstract

The animal genetics breeding represents a set of processes, implemented through its selection and mating tools, with the aim of maximizing its bioeconomic merit in the next generation of the herd. In this context, animal genetic breeding programs appear to help rural producers optimize means that lead to a greater economic return of their beef cattle activity, incorporating natural selection indices, and criteria that represent, for example, the economic return of the production. Inserting economic criteria into selection indices can facilitate increased profitability through decisions focused on a wider range of economically relevant traits. Considering mating as one of the fundamental tools of animal genetic improvement, this paper aims to implement a strategy to recommend the best bovine matings based on indices of selection by economic values. The optimization of this genetic improvement process involves the treatment of a variety of bioeconomic traits causing a combinatorial explosion of mating solutions. In computing, optimization problems are implemented by computational algorithms for their solution. In this approach it is presented a solution based on Evolutionary Algorithms (EA - Evolutionary Algorithm), that is, computational models, inspired by the process of natural evolution, for solving combinatorial optimization problems. In this context, the purpose of this paper is the application of a genetic algorithm to create mating recommendations for Brangus cattle taking into account the interesting traits to the Embrapa genetic breeding program. The data used in the simulations of this work correspond to the selection criteria: i) weight at birth; ii) mature cow weight; iii) longissimus muscle area; iv) backfat thickness; v) rump fat thickness; vi) flight speed; vii) number of nematode eggs per gram of faeces; and viii) tick count. The results confirm that the inference of the computational strategy for mating selection provides genetic evolution for the selected traits, expressing in economic values the gains in productivity.