Genetic evaluation to design a reference cow population for the Holstein breed in Tunisia: a first step toward genomic selection

Context There is a large gap between developed and developing countries in the field of genetic evaluation of livestock animals. Introduction of genomic evaluation procedures and methods could contribute to reducing this gap. Aims The goal of this study was to select a reference cow population for the Holstein breed in Tunisia. Specific objectives were to update adjustments for non-genetic factors affecting milk yield, estimate genetic parameters for milk yield, and predict cow breeding values. Methods A BLUP animal model was used for 11 175 lactations recorded between 2012 and 2017 from 6251 dairy cattle raised in 33 Holstein dairy herds from three types of herds or production sectors. A pedigree file of 16 211 males and females was included in the genetic evaluation. Multiplicative adjustment factors were computed for age and month of calving, using adjusted 305 days in milk. Key results Month of calving, age at calving and farm ownership were significant sources of variation for milk yield. Cows calving in autumn and early winter (September–January) yielded more milk than those calving in spring (February–May) by 430 kg, and summer (June–August) by 455 kg. Holstein cows in Tunisia reached their maximum milk yield during the fourth lactation. Average adjusted milk yield for days in milk, month and age of calving was 6621 ± 2883 kg. Heritability estimates of adjusted milk yield were 0.14 ± 0.02 for all seven lactations and 0.16 ± 0.03 for the three first lactations. Phenotypic correlations among lactations were all positive, ranging from 0.25 to 0.43. On the basis of two main traditional designs (extreme yield and top accuracy), 1000 cows were selected to form the Tunisian female reference population. Conclusion The first step of genomic evaluation has been realised by forming a reference population from cows selected for adjusted milk yield production, based on their predicted breeding values and accuracies, using a BLUP animal model. Implication In the absence of progeny testing and sufficient proven sires, a cow reference population could represent the alternative for implementing genomic selection in developing countries.