Are the tentative proxy traits themselves affected by CO2 level?
Linear correlations for the three tentative proxy traits measured under elevated vs. ambient CO2 levels were similar among the traits (Fig. 6A). Overall across the 17 genotypes, the slope was near 1, indicating the absence of a generic effect. However, with R² values between 0.5 and 0.6, there was considerable scatter, and the SEM error bars indicated that for some genotypes, there was an effect of CO2 level on the proxy trait value.
In fact, there generally was a clear negative trend (in some cases significant at P<0.05) in the relationship between the plasticity of the proxy traits (expressed as the difference between elevated and ambient CO2 levels) vs. the Amax or grain yield ratio observed between the CO2 levels (Fig. S1, supplemental materials). Consequently, those genotypes that had a reduced proxy trait value under e-CO2 due to phenotypic plasticity showed a smaller Amax or yield gain under e-CO2.
Correlations across genotypes of the other measured variables under elevated vs. ambient CO2 levels are shown in Fig. S2 (supplemental materials).