Parâmetros genéticos de características morfofuncionais e identificação de linhagens na raça Mangalarga Marchador

Abstract

The Mangalarga Marchador (MM) breed originating in the southern region in the state of Minas Gerais is the largest breed of horses raised in Brazil. The objectives of this study were to estimate the effects of the environment and the genetic parameters of linear characteristics and score scores (gait, morphology and total) obtained at the time of the definitive registration of MM animals, as well as to identify the existence of possible lines in this breed, taking into account the allele frequency observed from information obtained by genotyping in microsatellite markers using discriminant analysis of principal components. To obtain the genetic parameters, data from 167,764 animals from 14,252 farms between 1967 and 2016 were evaluated. The characteristics evaluated were heights: withers (HW) and croup (HC); lengths: head (LH), neck (LN), shoulder (LS), back-loin (LBL), croup (LC), body (LB); widths: head (WH) and croup (WC); perimeters: thoracic (PT) and cannon (PC); scores: gait (SG), morphology (Morf) and total (Stotal). The components of (co) variance, heritability (h²), genetic and phenotypic correlations between the linear characteristics were obtained using the restricted maximum likelihood method using the AIREMLF90 program. The characteristics obtained by scoring scores were assessed by Bayesian analysis, using the THRGIBBSF90 program and the POSTGIBBSF90 application. For the identification of the lines, data from 60,795 MM horses (24,037 males and 36,758 females) were used, distributed throughout the national territory and in all states of the federation, genotyped in a paternity test by DNA examination for 13 microsatellite loci between years 2005 to 2016. For verification of the possible existence of lines the Discriminant Analysis of the Principal Components (DAPC) was carried out, using the program R. The Bayesian Information Criterion - BIC was used to identify the number optimal number of genetic (K) groups. To choose the ideal number of main components (PCs) to be retained in the analysis, the α-score was calculated using the adegenet package of the program R. The h2 obtained were: HW (0.46); HC (0.46); LH (0.20), LN (0.21), LBL (0.27), LC (0.23), LS (0.21); LB (0.35); WH (0.24); WC (0.25); PT (0.22); PC (0.18); SG (0.21), Morf (0.22) and Stotal (0.26). The genetic correlations ranged from zero, between CP and And, to 0.96 between AC and AG. The phenotypic correlations ranged from zero, between AG and Morf, to 0.85 between And and Morf. The ideal number of groups was 300 and the main components retained in the DAPC were 35, chosen from the lowest BIC value (300) and highest α-score value (35 and 0.8), respectively. Was not observed the existence of dispersion of values along the DAPC graph, demonstrating the non-formation of groups or any genetic distance between animals within the MM breed. The effects of sex, farm, year of birth and registration must be included in the analysis model to estimate genetic parameters for the morphofunctional characteristics of the MM breed. Withers and croup height characteristics must respond to selection, resulting in genetic progress in MM animals. The selection for withers height should result in correlated changes in the height of the croup and in the body length, as well as the selection for total score may lead to the improvement of the genetic merit of the morphology of animals of the MM breed. The existence of lines in the MM breed was not genetically identified, considering data from microsatellites of animals evaluated from the year 2005.