Genetic Parameters of First Lactation Milk Yield Under Low, Medium and High Production Systems in Kenya, using Test-Day Random Regression Model

Abstract

The aim of this study was to estimate genetic parameters for test-day milk yield in different production systems in Kenya. 10,923, 19,049 and 26,287 first lactation test-day records from multiple breeds under low, medium and high production systems, respectively, were analysed. On average cows under high production systems were younger and had a higher test-day milk yield than in low and medium production systems. A model fitting fourth order Legendre polynomials was found to be the most parsimonious and was therefore used to model the data. Additive genetic and permanent environmental variances were heterogeneous along different days in milk and between production systems. Heritability and repeatability were also different between days in milk and production systems. Heritability was on average 27%, 48% and 48% and repeatability 72%, 83% and 78% under low, medium and high production systems, respectively. Genetic correlations ranged from -32%, 34% and 45% to unity between daily milk yield in different days in milk under low, medium and high production systems, respectively. These parameters indicate that random regression using Legendre polynomial order four can be used to model test-day milk yield under the three production systems in Kenya. The observed heterogeneity of variance indicates that genetic parameters should be estimated within production systems for sustainable genetic improvement.