A criticism of the value of midparent in polyploidization.

The hypothesis of genetic additivity states that the effects of different alleles, or different genes, add up to produce the phenotype. When considering the F1 progeny of a cross, the hypothesis of additivity of the genetic dosages provided by the parents is tested against the mid-parent value (MPV), which is the average of parental phenotypes and represents the reference value for genetic additivity. Non-additive effects (genetic interactions) are typically measured as deviations from MPV. Recently, however, the use of MPV has been directly transposed to the study of genetic additivity in newly synthesized plant polyploids, assuming that they should as well display mid-parent expression patterns for additive traits. It is shown here that this direct transposition is incorrect. It is suggested that, in neo-polyploids, mid-parent expression has to be reconsidered in terms of reduced genetic additivity. Homeostatic mechanisms are deemed to be the obvious ones responsible for this effect. Genomes are therefore ruled by negative epistasis, and heterosis in allopolyploids is due to a decreased interaction of the parental repressive systems. It is contended that focalizing on the right perspective has relevant theoretical consequences and makes the studies of neo-polyploids very important for our understanding of how genomes work.