Why were proxy traits for CO2 response more predictive when measured at elevated [CO2]?
For purposes of practical applications of LSSR in plant phenotyping and selection or breeding for rice CO2 response, it is sobering that in both our experiments the LSSR measured under elevated [CO2] was much more predictive than that measured under ambient [CO2]. (In Exp.1, the latter was not predictive at all, Fig. 1A.) Given our finding that genotypic LSSR was highly constitutive between CO2 levels, with a slope of 1 for the linear regression, we can exclude that this difference was caused by a common adaptive-plasticity response of the LSSR trait for the population as a whole. However, those genotypes that showed a reduction in LSSR (or values of alternative proxy traits discussed further down) under elevated [CO2] tended to have greater gain in A max and grain yield under elevated [CO2]. Although this trend was mostly not significant statistically, it is clear that variable phenotypic plasticity among the genotypes was responsible for the greater predictability of e-CO2 response with proxy traits measured under elevated [CO2].