Pathways for making unisexual flowers and unisexual plants:Moving beyond the 'two mutations linked on one chromosome' model

Sex determination in higher plants is of fundamental biological interest and has great practical significance for fruit yield and high-quality seed production. Th e last decades have witnessed an explosion in the genetic-developmental understanding of a typical hermaphroditic flower; however, the genes controlling dicliny (unisexual fl owers) and dioecy (unisexual individuals) are only now being revealed ( Akagi et al., 2014 ; Boualem et al., 2015 ). These genetic-developmental insights, from distantly related groups, now need to be brought together with the large body of empirical observations showing that dicliny and dioecy each evolved independently many thousands of times, but oft en in the same genera and families suggesting related evolutionary pathways ( Darwin, 1877 ; Lloyd, 1972 , 1975a , b , 1980 ; Renner and Ricklefs, 1995 ; Renner, 2014 ). Here I highlight how the recent insights may redirect our understanding of pathways to dioecy in fl owering plants. Th e fundamental model for the evolution of dioecy in angiosperms has long been that “at least two gene mutations are necessary to transform an hermaphroditic or monoecious species into one with separate sexes; one mutation must aff ect ovule production, and the other the production of pollen” ( Charlesworth and Charlesworth, 1978 : p. 975), with the corollary that “diff erent genes determining sex [are] grouped together in one region of one chromosome in most species” (p. 976). Th is model disregards that (1) in monoecious species, mutations suppressing male and female function do not need to arise since they already exist and that (2) there is little evidence that sex-determining genes in dioecious plants are grouped together on one chromosome. Th e developmental-genetic