Breeding strategies to make sheep farms resilient to uncertainty

The sheep industry in Western Australian has had many challenges over the last 20 years which have caused sheep numbers to decline. This decline is because sheep farms are not resilient to uncertain pasture growth and commodity prices. One way to improve resilience and profitability of farming systems is through breeding of sheep. Therefore, this thesis had two aims; 1. Quantify the potential to select and breed sheep that are more resilient and 2. Quantify how sheep breeding can make farming systems more resilient. To determine if sheep can be bred to be resilient to varying pasture growth I investigated if live weight change is a heritable trait. I investigated live weight change in adult Merino ewes managed in a Mediterranean climate in Katanning in Western Australia. Live weight change traits were during mating and lactation. The heritability of live weight change was low to moderate. Therefore that live weight change could be a potential indicator trait for resilience to uncertain pasture growth. To include live weight change in a breeding goal, correlations with other traits are needed. I calculated the genetic correlations between live weight change during mating, pregnancy and lactation, and reproduction traits. Most genetic correlations were not significant, but genetically gaining live weight during mating in two-year old ewes and during pregnancy for three-year-old ewes improved reproduction. Therefore, optimised selection strategies can select for live weight change and reproduction simultaneously. To investigate optimal breeding programs to make sheep farms resilient to uncertain pasture growth and prices, I modelled a sheep farm in a Mediterranean environment. The economic value of seven traits in the breeding objective were estimated. Including variation in pasture growth and commodity prices decreased average profit and increased the economic value of all breeding goal traits compared to the average scenario. Economic values increased most for traits that had increases in profit with the smallest impact on energy requirements. I also compared optimal breeding programs for across 11 years for 10 regions in Western Australia with different levels of reliability of pasture growth. I identified two potential breeding goals, one for regions with low or high pasture growth reliability and one for regions with medium reliability of pasture growth. Regions with low or high reliability of pasture growth had similar breeding goals because the relationship between economic values and reliability of pasture growth were not linear for some traits. Therefore, farmers can customise breeding goals depending on the reliability of pasture growth on their farm.