Do maize chromosomes contribute to genetic variance in proportion to their size in elite populations?

Quantitative trait loci (QTLs) are often assumed to be uniformly distributed across the genome in modeling and simulation studies. Our aim was to assess whether the contribution of each chromosome to genetic variance (V_G) is proportional to its length in maize (Zea mays L.). We estimated per-chromosome V_G from genomewide marker effects and linkage map positions in eight testcross populations, and we tested whether the slope of the regression of the observed V_G on expected V_G was different from b = 1. Our results indicated that a chromosome's contribution to V_G is generally proportional to its size, supporting the common model of QTLs being uniformly distributed across the genome. In particular, the b value was not statistically different from 1 in about 80% of the population trait combinations. However, 1-2 chromosomes often contributed disproportionately to V_G even when b did not differ from 1. The results indicated that these exceptions arose from the preponderance of either coupling linkages, which produced a higher-than-expected contribution, or repulsion linkages, which produced a lower-than-expected contribution. Because these exceptions resulted from combinations of alleles rather than the effects of alleles by themselves, maize breeding strategies will likely continue to focus on preserving or creating favorable combinations of existing QTL alleles rather than finding copies of individual chromosomes that are hotspots for QTLs with large effects.