Figure 1

Ideal vs. possible experimental designs. 1a) Ideal experimental design for each of the 4 plots containing 9 subplots of 27 shoots, with crossed levels of genotypic and allelic richness designed to disentangle their respective effects. b) Frequency distribution of allelic richness (in total number of alleles) across the three levels of genotypic richness (3 MLGs in light grey, 6 MLGs in medium grey and 9 MLGs in dark grey). c) Evolution of allelic richness for a broader range of levels of genotypic richness. d) Best possible design at the levels of genotypic richness manipulated in the experiments, illustrating the composition of each of the 4 plots made of 9 subplots of 27 shoots each, in terms of allelic richness with increasing levels of genotypic richness. The schemes of plot are designed with nested and increasing orders of richness for illustrative proposes, but the positions of the 36 sub-plots in the aquaculture tank where the experiment took place were randomized.