Assessing population structure and genetic diversity in U.S. Suffolk sheep to define a framework for genomic selection

Long-term sustainability of breeds depends on having sufficient genetic
diversity for adaptability to change, whether driven by climatic
conditions or by priorities in breeding programs. Genetic diversity in
Suffolk sheep in the U.S. was evaluated in four ways: 1) using genetic
relationships from pedigree data [(n=64,310 animals recorded in the U.S.
National Sheep Improvement Program (NSIP)]; 2) using molecular data (n=304
Suffolk genotyped with the OvineHD BeadChip); 3) comparing Australian
(n=109) and Irish (n=55) Suffolk sheep to those in the U.S. using
molecular data; and 4) assessing genetic relationships (connectedness)
among active Suffolk flocks (n=18) in NSIP. By characterizing genetic
diversity, a goal was to define the structure of a reference population
for use for genomic selection strategies in this breed. Pedigree-based
mean inbreeding level for the most recent year of available data was 5.5%.
Ten animals defined 22.8% of the current gene pool. The effective
population size (Ne) ranged from 27.5 to 244.2 based on pedigree and was
79.5 based on molecular data. Expected (HE) and observed (HO)
heterozygosity were 0.317 and 0.306, respectively. Model-based population
structure included 7 subpopulations. From Principal Component Analysis,
countries separated into distinct populations. Within the U.S. population,
flocks formed genetically disconnected clusters. A decline in genetic
diversity over time was observed from both pedigree and genomic-based
derived measures with evidence of population substructure as measured by
FST. Using these measures of genetic diversity, a framework for
establishing a genomic reference population in U.S. Suffolk sheep engaged
in NSIP was proposed.