An ABA-GA bistable switch can account for natural variation in the variability of Arabidopsis seed germination time.

And in fact, the QTL results also suggest that these phenotypes can't be uncoupled – any QTL for mode or percent is also a QTL for CV, and the QTL on chromosome 5 that the authors claim is CV-specific both explains little variation and its marginal significance depends on whether or not 8 particular lines with bimodal distributions (and therefore perhaps cases in which mode doesn't describe centrality well) are included in the analysis.

also in 5,069 other papers