Implications for TBMs
Accounting for diurnal dynamics of apparent Vcmaxand Jmax­ means that gross carbon assimilation could be mis-estimated ~-4 to 40% (standard deviation based on the average and peak Vcmax scenarios relative to the diurnal scenario). If extrapolated to global gross primary production (~123 Pg C yr­-1; Beer et al ., 2010), diurnal variation in photosynthetic parameters could create a revised estimate range of 101 ± 27 Pg C yr-1 (mean ± standard deviation) in global GPP. This uncertainty introduced here is similar to the uncertainty in unconstrained CMIP5 estimates of GPP (standard deviation: 27.5 Pg C yr-1; Mystakidis et al ., 2016), implying that understanding diurnal processes regulating Vcmaxmay warrant more attention. Although limited in scope, this study clearly shows that diurnal variation in apparentVcmax and Jmax exists across multiple species under greenhouse and field conditions. While there are factors such as mesophyll conductance,Kleaf , and Rubisco activation state that could cause these diurnal dynamics in apparent Vcmaxsuch that there is no effect on the amount of Rubisco in the leaf, it is important to note that TBMs currently use apparentVcmax to model carbon assimilation (Rogerset al ., 2017). TBMs are also not ready for incorporation ofgm , Rubisco activation state, andKleaf due to a paucity of data for these parameters. Therefore, if multiple processes are producing diurnal changes in apparent Vcmax , then including diurnal dynamics in apparent Vcmax may present a relatively simple way to include these important dynamics in TBMs. However, this requires broad diurnal RACiR surveys, paired with measurements such as gm , Rubisco activation state, and Kleaf , to confirm the underlying mechanisms behind diurnal Vcmax .
Implementing diurnal changes in Vcmax into TBMs requires more physiological understanding of this phenomenon. Our data provide preliminary evidence that diurnal variation inVcmax may depend on the evolutionary history of the organism as well (the two species from Pinaceae did not show a diurnal pattern), however this could be related to different diurnal dynamics (or lack thereof) in gm , Rubisco activation state, and Kleaf . These areas still require a marked advance in process knowledge ofgm (Rogers et al ., 2017), Rubisco activation state, and Kleaf before they could be implemented into TBMs. However, ignoring these diurnal dynamics could lead to a mis-estimation of CO2 assimilation by TBMs, especially since diurnal variation could be as significant as other sources of variation (e.g. leaf age, Wu et al ., 2016; species, Wullschleger, 1993; model structure, Mystakidis et al ., 2016). Greater mechanistic understanding of diurnal variation inVcmax and quantification of the phenomenon in key biomes is required to so that improved model formulations of photosynthesis can be considered for inclusion in future TBMs.