The photorespiratory CO 2 compensation point (Γ*) and the rate of CO 2 release in the light (D L) are critical parameters for understanding the carbon dynamics of C 3 plants. These two parameters can be derived from the widely-used Laisk method as the intercept of linear regression lines fitted to net assimilation rates (A net) versus chloroplast CO 2 partial pressures (C c) obtained at different low-irradiance levels. However, photosynthetic theory indicates curvature in the A net-C c relationship which conflicts with the use of linear relationships for analysis. We, therefore, systematically evaluated the limitations of the use of linear relationships across temperatures from 5 to 40 °C and quantified the sensitivity of errors in Γ* and D L estimates to the selected C c range. We found that wide CO 2 ranges, especially when they exclude the actual Γ*, can introduce substantial biases in parameter estimation with linear regressions, particularly at lower temperatures. It can lead to marked underestimates of Γ*, and biologically unrealistic D L. We propose refining the Laisk method by using a photosynthesis model to analyse data. The model better represents the non-linear A net-C c relationship and yields consistent Γ* and D L estimates, regardless of the CO 2 range used.
