Including second lactation data in Canadian feed efficiency evaluation

Abstract

The Canadian 1^st lactation genomic evaluation model for feed efficiency was expanded by inclusion of 2^nd lactation phenotypes, which account for 30% of all available international data. The new model is a 12-trait (lactation by DIM interval by trait) animal model, with assumptions following the current official implementation. Feed Efficiency (FE) is defined as genetic Residual Feed Intake in 61- 305 DIM within each lactation, with Energy Corrected Milk and Metabolic Body Weight as energy sinks for Dry Matter Intake (DMI). Estimates of heritability for FE were 0.06 and 0.03 in 1^st and 2^nd lactation, respectively. Lactations were weakly genetically correlated (r_g = 0.25) for FE. Each sink was genetically uncorrelated with FE, by definition within each lactation, and genetic correlations between FE and DMI were around 0.4. There were 8,927 cows with DMI data in December 2021 test-run, of which 81% were genotyped. Inclusion of 2^nd lactation data increased the Single-Step reference population size (N = 12,942) by 17%, compared with the 1^st lactation only model. Average reliability of FE for a sample of 50,000 young, genotyped animals born in 2021 was around 50% for both within lactation traits and the FE Index, defined as an average of 1^st and 2^nd lactation FE. Reliability of 1^st lactation FE was larger, on average, by 1.5 points compared with estimates using only 1^st lactation data. Correlations between FE Index and within lactation FE were larger than 0.8. First lactation FE correlated well (r = 0.97) with the current official evaluations. A 5-point increase in RBV (mean = 100, SD = 5 for base bulls, reversed in sign) for FE Index is expected to result in a reduction in feed intake of bull’s daughters by approximately 2%. The new Canadian FE model is scheduled for implementation in December 2022, with a simultaneous inclusion of FE in LPI and Pro$ indices.