Modeling Anet
We modelled photosynthesis according to Farquhar et al . (1980), where Anet is the minimum of carboxylation and electron transport limitations:
\(A_{\text{net}}=min\left(A_{c},\ A_{j},\ A_{t}\right)-R_{d}\)(Equation 1)
\(A_{c}=V_{c,max}\frac{C_{i}-\Gamma^{*}}{C_{i}+K_{c}\left(1+\frac{O}{K_{o}}\right)}\)(Equation 2)
\(A_{j}=J\frac{\left(C_{i}-\Gamma^{*}\right)}{\left(4C_{i}+8\Gamma^{*}\right)}\)(Equation 3)
\(\left[J-0.5\left(1-f\right)I\right]\left[J-J_{\max}\right]=0\)(Equation 4)
\(A_{t}=V_{c,max}/2\) (Equation 5)
Where Ac, Aj,, andAt are carboxylation, electron transport, and triose phosphate utilization limited photosynthesis, respectively;Rd is respiration in the light;Vc,max is the maximum rate of Rubisco carboxylation, Ci is leaf intercellular [CO2]; Γ* is the CO2 compensation point in the absence of mitochondrial respiration; Kc and Ko are the Michaelis-Menten constants for Rubisco carboxylation and oxygenation, respectively; O is the O2concentration in the chloroplast (O = 210 μmol mol-1); J is the rate of photosynthetic electron transport; f is the fraction of light energy not absorbed by the chloroplast (0.15); I is the incident irradiance; and Jmax is the maximum rate of photosynthetic electron transport.
Vc,max , Jmax ,Kc , Ko , and Γ* were thermally scaled using an Arrhenius equation:
\(f\left(T\right)=k_{\text{ref}}\exp\left[E_{a}\frac{T-T_{\text{ref}}}{T\times T_{\text{ref}}\times R_{g}}\right]\)(Equation 6)
Where f(T) is the parameter at temperature T in K;kref is the parameter at the reference temperature (either 25 °C or 30 °C); Tref is the reference temperature in K (either 298 K or 303 K);Ea is the activation energy in kJ mol-1; and Rg is the universal gas constant (0.008314 kJ K-1mol-1). See Bernacchi et al . (2001) forEa , and kref for all parameters except Jmax , and see Bernacchiet al . (2003) for Ea forJmax (note that kref forVc,max and Jmax were derived from measured RACiR curves for each species).
Respiration (R ) was modelled as according to Atkin & Tjoelker (2003):
\(f\left(T\right)=10^{\frac{{T-T}_{\text{ref}}}{10}\times\operatorname{}{Q_{10}+\operatorname{}R_{\text{ref}}}}\)(Equation 7)
Where f(T) is the parameter at temperature T in K;Q10 is the thermal sensitivity coefficient (assumed to be 2); and Rref is respiration at the reference temperature.
Since the photosynthetic rates depend on Ci , we modelled CO2 diffusion into the leaf using Moss and Rawlins (1963):
\(A_{\text{net}}=g_{s}\left(C_{a}-C_{i}\right)\) (Equation 8)
Where gs is stomatal conductance (mol m-2 s-1), Caand Ci are the concentrations of CO2 at the leaf surface and intercellular airspace, respectively (μmol mol-1).
To close the system of equations, we used the stomatal conductance model from Medlyn et al . (2011) as implemented by Lin et al . (2015):
\(g_{s}=1.6\left(1+\frac{g_{1}}{\sqrt{D}}\right)\frac{A_{\text{net}}}{C_{a}}\)(Equation 9)
Where g1 is the model coefficient equal to 4.16, and D is vapor pressure deficit.
Daily net leaf carbon uptake was modelled in R assuming an ambient CO2 concentration (Ca ) of 400 μmol mol-1, and constant dark respiration (R ) at 25 °C (unless otherwise stated) (2.27 ± 0.16 µmol m-2 s-1 for Populus deltoides , 0.46 ± 0.15 µmol m-2 s-1for Capsicum annuum , 3.36 µmol m-2s-1 for Malus domestica (Bunce, 1992), 2.30 ± 0.16 µmol m-2 s-1 for Populus trichocarpa , 2.49 ± 1.00 µmol m-2s-1 for Asclepias speciosa (30 °C), 0.74 ± 0.07 µmol m-2 s-1 for Quercus muehlenbergii (30 °C), 1.85 ± 1.08 µmol m-2s-1 for Rosa grandiflora (30 °C), 0.24 µmol m-2 s-1 for Pinus ponderosa(10 °C; Law et al ., 2001), 0.71 ± 0.11 µmol m-2s-1 for Ginkgo biloba (30 °C), and 0.85 ± 0.22 µmol m-2 s-1 for Cedrus deodar (30 °C)). Environmental conditions at the University of New Mexico were taken for June 28, August 24, and October 9 in 2018 (Fig. 1) and used to provide driving data for all species. Half-hourly time intervals were summed for total carbon uptake for all species. Models were run for four scenarios (1) Vcmax andJmax scaled from maximumVcmax and Jmax to account for diurnal dynamics, (2) maximum measured Vcmaxand Jmax , (3) minimum measuredVcmax and Jmax , and (4) average measured Vcmax andJmax . All R code is available as a supplementary file (“Stinziano et al Modeling.Rmd”, “environmentdata.csv”). We also tested the diurnal modeling by modeling the leaf chamber conditions used to measure RACiR at each time point for each species, comparing modelled Anet and gs at 400 μmol CO2 mol-1. We then tested model performance by using the {lm} function in R (R Core Team, 2018).