The genomic impacts of drift and selection for hybrid performance in maize

Authors: GERKE, JUSTIN - University Of Missouri; Edwards, Jode; Guill, Katherine; ROSS-IBARRA, JEFFREY - University Of California; McMullen, Michael

Submitted to: Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/11/2015
Publication Date: 9/18/2015
Publication URL: http://handle.nal.usda.gov/10113/62260
Citation: Gerke, J.P., Edwards, J.W., Guill, K.E., Ross-Ibarra, J., McMullen, M.D. 2015. The genomic impacts of drift and selection for hybrid performance in maize. Genetics. 201(1201-1211). doi: 10.1534/genetics.115.182410.

Interpretive Summary: Corn production in the United States is based on growing hybrid lines. The hybrid is formed by crossing two divergent inbred lines. For efficient development of improved hybrids corn breeders maintain inbreds in genetically isolated groups of germplasm known as heterotic pools , with a hybrid resulting from lines crossed from different pools enhancing hybrid vigor. Little is known about the effects on genetic diversity and genome structure that result from this current breeding paradigm. In this study we determined the genetic changes resulting from a 16 generation breeding program. We documented a loss of genetic diversity largely consistent with random effects known as genetic drift. However, this study also provides evidence for selection in specific genomic regions and for an important interplay between local recombination rate (the rate of exchange of chromosomal segments) and efficacy of selection. This study is important because it illuminates limitations placed on maize improvement caused by the low recombination in a large fraction of the genome. This study indicates the need to understand and potentially modify recombination rate and selection strategies to speed corn improvement.

Technical Abstract: Modern maize breeding relies upon selection in inbreeding populations to improve performance in cross-population hybrids. The United States Department of Agriculture - Agricultural Research Service (USDA-ARS) reciprocal recurrent selection experiment between the BSSS and BSCB1 populations represents one of the longest standing models of selection for this type of hybrid performance. To investigate the genomic impact of this selection program, we used the Illumina MaizeSNP50 high-density SNP array to determine genotypes of progenitor lines and over 600 individuals across multiple cycles of selection. Consistent with previous research, we found that genetic diversity within each population steadily decreases, with a corresponding increase in population structure. High marker density also enabled the first view of haplotype ancestry, fixation and recombination within this historical maize experiment. Extensive regions of haplotype fixation within each population are visible in the pericentromeric regions, where large blocks trace back to single founder inbreds. Simulation attributes most of the observed reduction in genetic diversity to genetic drift. Signatures of selection were difficult to observe in the background of this strong genetic drift, but heterozygosity in each population has fallen more than expected. As observed previously, the regions most likely targeted by selection do not overlap between the two populations. We discuss how this pattern is likely to occur during selection for hybrid performance, and how it poses challenges for dissecting the impacts of modern breeding and selection on the maize genome.