2 Relationships of transpiration efficiency (TE) and leaf carbon isotope composition (δ13C) among 96 natural accessions of Arabidopsis thaliana. Symbols represent best linear unbiased predictors (BLUPs) associated with breeding values for each accession (see text). Open and filled symbols represent spring and winter accession means, respectively. Lines represent linear regression; r 2 and P values are given Variation in components of WUE The CP673451 nmr 18 natural accessions of Arabidopsis in experiment 2 were selected to represent a wide range of intrinsic WUE as indicated by δ13C (Table 1). Whole-plant gas exchange measurements

in a custom cuvette (Fig. 1) OICR-9429 mouse showed that these lines also exhibit considerable variation in whole rosette A and g s in a common environment (Fig. 3). Accession mean whole rosette A ranged between 10 and 16 μmol m−2 s−1, but the heritability was not significantly different from zero (P = 0.137). g s showed significant genetic variation, ranging between 0.17 and 0.45 mol m−2 s−1 with a heritability of H 2 = 0.33 (accession P value = 0.002). In addition, g s was a better predictor of variation in δ13C than A. We found a significant negative correlation between δ13C and g s among accessions (r 2 = 0.40, P = 0.0027), and a weaker correlation between δ13C and A (r 2 = 0.25,

P = 0.036). In general, the high conductance check details lines had low intrinsic WUE, as indicated by δ13C, but there was a wide range of δ13C in the MG132 low conductance lines, suggesting additional sources of variation. The expected negative correlation between δ13C and g s was largely caused by the spring accessions. The winter accessions tended to show the opposite pattern (not significant), with the exception of Tamm-2, an accession from Finland that had the highest g s

of all. Fig. 3 Relationships between assimilation (A), stomatal conductance (g s), and leaf carbon isotope composition (δ13C) at 350 μmol photons m−2 s−1 from whole-shoot gas exchange of 18 accessions of Arabidopsis selected from the larger panel of accessions to represent extremes in δ13C. Open and filled symbols represent spring and winter accession means, respectively. Lines represent linear regression; r 2 and P values are given Despite the lack of heritability of A and the weak correlation of A with δ13C, we did find a significant positive correlation between g s and A among accessions (r 2 = 0.78, P = 0.00001). This is consistent with the optimization of stomatal regulation to maximize carbon gain while minimizing the water loss (Katul et al. 2010). Accessions that have high conductance should be under selection for increased biochemical capacity (Bloom et al. 1985). Although, it is not formally stated, such optimality approaches interpret consistent patterns of correlation in physiological traits (Reich et al.