• Optimal system of contract matings for use in a commercial dairy population

      McParland, Sinead; Kearney, K. F.; Lopez-Villalobos, N.; Berry, Donagh (Teagasc (Agriculture and Food Development Authority), Ireland, 2009)
      Managing the contribution of prominent animals to the pedigree of livestock populations is a topic of increasing importance worldwide. The aim of this study was to evaluate methods of controlling the accumulation of inbreeding in the Irish Holstein-Friesian population through the methodology used to arrange contract matings. Two non-random mating systems were investigated, linear programming (LP) and sequential programming (SEQ); these were compared with random mating (RAN) and mating of the best sires to the best dams (TOP). All mating systems were compared across a range of objectives: to maximise genetic merit for the economic breeding index (EBI) used in Ireland, to minimise population coancestry with breeding females (R-value), and a dual objective of simultaneously maximising EBI and minimising coancestry with breeding females. Algorithms were developed to identify elite dams and sires from the national herd for use in the contract mating programme. One thousand contract matings were generated using each selection method, with the aim of producing 83 test sires (the number of bulls which it is feasible to test annually in Ireland) for use in a progeny testing scheme. The top 1,000 matings, as selected by the LP and SEQ methods, performed similarly when maximising the dual objective (average progeny EBI of €145 and an average coancestry of the progeny to the population of breeding females of 0.93%). The TOP and RAN methods both selected phantom progeny with higher coancestry with the breeding female population (1.21% and 1.34%, respectively) than the LP and SEQ methods. However the matings from the TOP method generated progeny of higher genetic merit (EBI = €199), whilst the progeny generated from the RAN method had lower genetic merit (EBI = €127) than those selected by the LP or SEQ methods.