Measurement and modeling bias assessment
To determine whether repeated clamping could bias the measurements (i.e.
through leaf damage), we ran RACiRs after 1 and 5 one-minute clamps on
the three greenhouse species (Capsicum annuum , Populus
deltoides , and Populus trichocarpa ). We found no significant
decline between clamps 1 and 5 for Vcmax(Capsicum annuum : t2 = -1.47, P =
0.28; Populus deltoides : t4 = -1.82,P = 0.14; Populus trichocarpa : t5 =
-1.67, P = 0.16; Fig. 6a), Jmax(Capsicum annuum : t2 = -1.43, P = 0.29;Populus deltoides : t4 = -2.01, P =
0.11; Populus trichocarpa : t5 =
-1.70, P = 0.15; Fig. 6b), and Anet(Capsicum annuum : t2 = 1.39, P =
0.40; Populus deltoides : t4 = -1.84,P = 0.21; Populus trichocarpa : t2 =
0.48, P = 0.68; Fig. 6c).
To check for model biases, we compared model predictions of the
leaf-level gas exchange to measured Anet andgs . The model consistently over-estimatedAnet by 9.8% (slope = 1.098,R2 = 0.96, F1,63 = 1356,P < 0.0001) while underestimatinggs by ~18% (slope = 0.82,R2 = 0.85, F1,63 = 345,P < 0.0001) (Fig. 7).