Prediction of the growth response to elevated CO 2 : a search for physiological criteria in closely related grass species.

Using 11 closely related grass species, we tested the capacity of physiological criteria to predict the growth response to elevated CO ₂ and to categorize the species with regard to their CO ₂ response. A growth analysis was conducted under productive conditions both at ambient (350 μmol mol ^-1 ) and elevated (700 μmol mol ^-1 ) CO ₂ . The relative growth rate stimulation was regressed against each of the growth rate components measured at ambient CO ₂ . Growth response to CO ₂ was positively correlated with specific leaf area (sla, the leaf surface area per unit of leaf weight), leaf area ratio (the leaf area per unit of total plant dry weight) and negatively correlated with net assimilation rate and leaf nitrogen concentration, both per unit of leaf area. We suggest that sla has a predominant role in these relationships. Different hypotheses are proposed and discussed in order to explain why species with low sla are less responsive to elevated CO ₂ . Neither biomass allocation, relative growth rate, shoot or root specific activities per unit of mass, nor chemical composition were significantly correlated with growth response to CO ₂ . The four predictive criteria mentioned above coherently differentiate the five wild annual species (higher sla, stronger growth response to CO ₂ ) from the four wild perennials. The two perennial crop species, with the highest sla, were more responsive than the wild species.