The validity of optimal leaf traits modelled on environmental conditions.

The ratio of leaf intercellular to ambient CO2 (chi) is modulated by stomatal conductance (g(s)). These quantities link carbon (C) assimilation with transpiration, and along with photosynthetic capacities (V-cmax and J(max)) are required to model terrestrial C uptake. We use optimization criteria based on the growth environment to generate predicted values of photosynthetic and water-use efficiency traits and test these against a unique dataset. Leaf gas-exchange parameters and carbon isotope discrimination were analysed in relation to local climate across a continental network of study sites. Sun-exposed leaves of 50 species at seven sites were measured in contrasting seasons. Values of chi predicted from growth temperature and vapour pressure deficit were closely correlated to ratios derived from C isotope (delta C-13) measurements. Correlations were stronger in the growing season. Predicted values of photosynthetic traits, including carboxylation capacity (V-cmax), derived from delta C-13, growth temperature and solar radiation, showed meaningful agreement with inferred values derived from gas-exchange measurements. Between-site differences in water-use efficiency were, however, only weakly linked to the plant's growth environment and did not show seasonal variation. These results support the general hypothesis that many key parameters required by Earth system models are adaptive and predictable from plants' growth environments.