Background/Question/Methods
Current models of carbon flux treat respiration simplistically.  As a first step towards a process-based representation, we tackled several salient questions about leaf respiration: how to represent the relative supply and demand for energy, reductant, and biosynthetic products, and how these are co-determined by respiratory and photosynthetic carbon and energy flows.  For example, the partial suppression of respiratory CO₂ release (R_c) in the light may involve changes in photoreductant supply and demand, photorespiratory NADH production, and anabolic carbon flows, and in the coupling among these processes.  Similarly, differences in R_c and its temperature dependency among tissue types and ages, and among species, likely result from differences in redox states of anabolic products.  We present a model designed to generate formal, testable hypotheses about the stoichiometric and regulatory basis of these processes.  The model assumes that flows of reductant, energy and carbon are regulated to satisfy specified demands, subject to flexible hypotheses concerning the degree of coupling among subcellular reductant pools.  The model is analytically soluble and is coupled to the photosynthesis model of Farquhar, von Caemmerer and Berry (1980). 
Results/Conclusions
The model predicts that suppression (S) of R_c in the light (S=100*(1-R_d/R_n), where R_d and R_n are R_c in the light and dark, respectively) is generally greater in mature leaves than in fast-growing leaves.  The value of S depends on assumed values for the fraction of excess photoreductant exported to the cytosol (f_x) and the fraction of photorespiratory NADH (produced by glycine decarboxylation) that remains in mitochondria (f_m) rather than being exported to peroxisomes.  For high f_m and/or f_x, S tends to be higher, because high reductant supply to the mitochondrial electron transport chain offsets the need for substrate oxidation to meet ATP demands (e.g., for mature leaves with f_m=0.6, S=38% if f_x=0 and 96% if f_x=1).  Suppression of R_c in the light in the model is stoichiometrically coupled to engagement of alternative electron sinks, which is predicted to increase with light.  The model also predicts a Kok effect at very low light due to suppression of oxidative pentose phosphate pathway activity; this effect is greater in young leaves due to their greater anabolic reductant demand. 
We conclude that it may be possible to predict R_c and S analytically, from stoichiometric constraints, given (i) knowledge of anabolic demands, including diurnal variation, and (ii) improved knowledge of the magnitude and regulation of reductant transport among subcellular reductant pools.