Count Bayesian models for genetic analysis of in vitro embryo production traits in Guzera cattle

Four models for in vitro embryo production traits in Guzera cattle were compared: Gaussian (untransformed variable - LIN and transformed in logarithmic scale LOG), Poisson (P01) and zero-inflated Poisson (ZIP). Data consisted of 5716 ovum pick-up and in vitro fertilization records performed in 1205 cows from distinct regions of Brazil. Analyzed count traits were the number of viable oocytes (Nov), number of grade I oocytes (N-GI,), number of degenerated oocytes (N-DG), number of cleaved embryos (N-CLV) and number of viable produced embryos (N-EMB). Heritability varied from 0.17 (LIN) to 0.25 (P01) for Nov; 0.08 (LOG) to 0.18 (ZIP) for N-GI; 0.12 (LIN) to 0.20 (P01) for N-DG; 0.13 (LIN) to 0.19 (P01) for N-av; 0.10 (LIN) to 0.20 (P01) for NEMB depending on the considered model. The estimated repeatability varied from 0.53 (LOG) to 0.63 (P01) for Nov; 0.22 (LOG) to 0.39 (ZIP) for No; 0.29 (LIN) to 0.42 (ZIP) for N-DG; 0.42 (LIN) to 0.59 (P01) for N-av; 0.36 (LIN) to 0.51 (P01) for NEMB. The goodness of fit, measured by deviance information criterion and mean squared residuals, suggested superiority of POI and ZIP over Gaussian models. Estimated breeding values (EBV) obtained by different models were highly correlated, varying from 0.92 for Nov (between LIN-P01) and 0.99 for N-GI (between POI-ZIP). The number of coincident animals on the 10% top EBV showed lower similarities. We recommend POI and ZIP models as the most adequate for genetic analysis of in vitro embryo production traits in Guzera cattle.