Rates of Genetic Gain for Alternative Indigenous Chicken

In-silico predictions to project the response to selection in livestock breeding programmes are done to estimate the possible genetic and economic gains. In these predictions, genetic and phenotypic parameters from various studies are used as some of the input parameters. Since parameters are affected by various factors such as sample sizes and genetic gain is a function of these parameters, the sources from which these parameters are obtained should be accounted for. Reduction of genetic variance due to selection (Bulmer effect) should also be accounted for since it reduces long-term genetic gain. Genomic selection should also be considered as it can accelerate the rate of genetic gain. There is therefore a need to consider the effect of sources of information, Bulmer effect and potential benefits of genomic selection on response to selection in indigenous chicken (IC) breeding programmes. The objective of this study was to contribute to genetic improvement of IC through estimation of pooled parameter estimates, accounting for Bulmer effect in response to selection and integration of genomic selection in IC breeding programme. The study addressed three specific areas; 1) Estimation ofpooled parameter estimates for traits of economic importance in IC breeding programme2) Estimation of Bulmer equilibrium genetic gains for a closed nucleus IC breeding programme and 3) Comparison of genetic gains for an IC breeding programme utilising genomic and conventional selection.Metaanalysis was used to compute pooled parameter estimates while deterministic computer simulation approach was used to model, estimate and evaluate Bulmer effect and response to selection.Two conventional strategies utilising pooled (CSP) and non-pooled parameters (CSN) and one genomic selection (GSS) strategies were considered in the simulation. The results demonstrate that parameters obtained through meta-analysis deviated from the ones obtained from single studies. The deviation for heritability for body weight at twenty weeks was 0.23 from the pooled value. The response to selection for CSN was 1.5 times more than response realised in CSP. The loss in genetic variances in CSN was 38% lower than that obtained in CSP. The GSS realised additional 54.5% and 60% accuracy and response to selection, respectively compared to CSP. Genomic selection had a reduced rate of inbreeding by 67.6% compared to CSP. It is concluded that use of non-pooled parameter estimates leads to over-estimation of potential response to selection and therefore pooled parameters should be used in modelling animal breeding programmes. It is also concluded that genomic selection optimises genetic gains and reduce rates of inbreeding in IC breeding programmes