Laisk measurements in the nonsteady state: Tests in plants exposed to warming and variable CO2 concentrations

Nonsteady-state Laisk measurements provide similar estimates of light respiration to steady-state methods but can be completed 4 times faster.


Supplemental Figure
. Respiration in the light estimated according to the DAT and the steady-state Laisk methods (LaiskDAT or LaiskSS; respectively) on paper birch (Betula papyrifera) from each of the six environmental treatments.(a) Area-based respiration in the light (RL-area) and (b) per nitrogen-based respiration in the light (RL-N).Colors represent the temperature treatments, with ambient temperature (T0) in blue, ambient +4°C (T4) in purple, and ambient +8°C (T8) in red.Hatching denotes measurement method (either hatched to represent LaiskDAT, or not, representing LaiskSS).The boxplots represent the median as well as the first and third quartiles.Whiskers delimit the range for each group, with outliers falling outside 1.5 x the interquartile range marked by points.Significance of the main effects for each trait as determined by a three-way ANOVA are noted (*** p<0.001; ** p<0.01).Full ANOVA results are found in Table S1.Different letters denote significant pairwise differences (p<0.05) as determined via a Tukey post-hoc comparison.n = 5 -6.S4. 1:1 comparison of RL-area calculated according to the DAT and the steady-state (SS) Laisk methods on paper birch (Betula papyrifera).Colors represent the temperature treatments, with ambient temperature (T0) in blue, ambient +4°C (T4) in purple, and ambient +8°C (T8) in red.Circles denote the ambient CO2 treatments (AC) and triangles the elevated CO2 treatments (EC).The dashed line shows the 1:1 line.n = 5 -6.S1.Summary of the three-way ANOVA in paper birch (B.papyrifera).Fvalues and p-values with temperature, CO2 and measurement method as the main effects are shown.Traits analyzed were area-and per nitrogen-based respiration in the light (RL-area and RL-N, respectively).Bold numbers represent p<0.05.), slope parameter estimates (Slope), 95% confidence intervals (95% CI), test of whether the slope equals 1 (H0 #1 Slope = 1), intercept parameter estimates (Intercept), and test of whether the intercept equals 0 (H0 #2 Intercept = 0).

Supplemental Text S1. Graphical methods for determining respiration in the light and the rubisco compensation point
The value for Γ* (rubisco compensation point) that was modeled from the gas exchange data varied and pre-treating leaves with high or low CO2 could create a difference of >0.7 Pa .In theory, Γ* is that CO2 concentration at rubisco at which the velocity of carboxylation is exactly ½ of the oxygenation velocity, assuming that one CO2 is released for every two oxygenation events.The rubisco compensation point should be a function of rubisco kinetics and should not vary from one plant to another of the same species nor vary with growth conditions.We consider here two possible effects that could account for the variability of Γ*.

Mesophyll conductance
It is not possible to measure Γ* directly but instead Ci* is measured, this is the CO2 concentration in the air of the intercellular space of the leaf at which the velocity of carboxylation is exactly ½ of the oxygenation velocity.The difference from Γ* is the drop in CO2 caused by respiration in the light (RL) and diffusion resistance of the cell, which is typically reported as the inverse, mesophyll conductance, gm.The relationship between Ci* and Γ* is Often, A, the CO2 assimilation rate, is used instead of -RL but because rubisco is, in theory, at its compensation point A = -RL.Thus, variations in observed Ci* should be a function of RL / gm and the computed Γ* should be constant for a given species.If RL= 0.5 μmol m -2 s -1 , then gm would cause Ci* to vary as shown in Figure 1.If gm is greater than 1 μmol m -2 s -1 Pa -1 , it would be possible that errors in the determination of gm could cause an error in Γ* determination of 0.5 Pa (Ci* value at infinite gm -Ci* value at a gm of 1 μmol m -2 s -1 Pa -1 ).We consider it unlikely that estimates of gm were inaccurate enough to account for the apparent variation in Γ*, especially because graphical methods for gm determination become more robust at low values of gm.

Glycine export from photorespiration
Carbon can leave the photorespiratory pathway.Under some conditions, this can lead to reverse sensitivity of photosynthesis to CO2 and O2 (Harley & Sharkey 1991).This has been modeled in detail by Busch et al. (2018) Equation 2 where αG is the amount of carbon that leaves photorespiratory metabolism as glycine and αS is the amount that leaves as serine.ΦG is (1-αG)⋅2⋅Γ*/C.right panel).The difference between the effect of αG and αS is because when glycine is exported, the CO2-liberating step is bypassed but not when serine is exported.The effect of just 10% of the carbon in glycine leaving photorespiratory metabolism is similar to the effect on apparent Γ* seen in response to pre-treating with either high or low CO2.
In neither case was the estimation of RL affected.

Figure 1 .
Figure 1.Variation in C i * as affected by g m .Assumptions include R L = 0.5 μmol m -2 s -1 .

Figure 2 .
Figure 2. Modeled photosynthesis rates using Equation2.R L was 0.5 μmol m -2 s -1 , Γ* was 4 Pa, and g m was considered infinite, i.e. not used in this modeling.