Table 1. Parameters derived from C_i response curves of A_n.

	Vcmax	Jmax	TPU	RMSE
Col-0	53 ± 1	100 ± 2	7.1 ± 0.1	0.93 ± 0.04
rca-2	40 ± 1***	86 ± 2***	6.7 ± 0.1 n.s.	0.95 ± 0.11 n.s.
rwt43	57 ± 3 n.s.	105 ± 5 n.s.	7.5 ± 0.2 n.s.	0.98 ± 0.07 n.s.
spsa1	54 ± 4 n.s.	86 ± 5**	5.5 ± 0.3***	0.85 ± 0.06 n.s.
aba2-1	58 ± 3 n.s.	118 ± 6***	8.5 ± 0.6**	1.12 ± 0.11 n.s.
C24	44 ± 2**	79 ± 4***	5.5 ± 0.4***	0.76 ± 0.07*
npq1-2	52 ± 3 n.s.	101 ± 5 n.s.	7.4 ± 0.4 n.s.	0.95 ± 0.08 n.s.
npq4-1	53 ± 1 n.s.	92 ± 2*	6.8 ± 0.2 n.s.	0.98 ± 0.03 n.s.

V_cmax, maximum caboxylation rate by Rubisco (μmol CO₂ m⁻² s⁻¹); J_max, maximum rate of electron transport in the absence of regulation (μmol electrons m⁻² s⁻¹); TPU, maximum rate of triose phosphate utilisation (μmol CO₂ m⁻² s⁻¹). The root mean squared error (RMSE, μmol CO₂ m⁻² s⁻¹) of the differences between measurement and model during curve fitting⁵⁵ is shown as an estimation of the overall goodness of fit. Averages ± SEM, n = 5–15. Stars within columns denote significance levels compared to Col-0: ***P < 0.0001, **P < 0.01, *P < 0.05. Absence of stars denotes lack of significant difference with Col-0 (P > 0.05).