Adapting an optimality-based model to predict half-hourly carbon uptake by ecosystems

Carbon dioxide (CO₂) uptake by leaves and its conversion into sugar by photosynthesis – gross primary production (GPP) – is the basis for vegetation growth. GPP is important for the carbon cycle, and its interactions with climate are a subject of study in Earth System modelling. One assumption of many current ecosystem models is that key photosynthetic traits, such as the capacities for carboxylation (V_cmax) and electron transport (J_max­) for ribulose-1,5-bisphosphate (RuBP) regeneration, are constant in time for any given plant functional type. Optimality theory predicts they should vary systematically with growth conditions, both in space and in time, and are not necessarily depend on the plant functional type. Moreover, theory makes specific, quantitative predictions about their (acclimated) community-mean values, predictions well supported by evidence. Neglecting such acclimation could lead to incorrect model estimates of the responses of primary production to climate change.

We focus on a proof-of-concept based on a primary production model, the P-model – which combines the Farquhar-von Caemmerer-Berry model for C₃ photosynthesis with eco-evolutionary optimality principles for the co-optimization of carboxylation and water transport costs – to allow the model to reproduce short-term variations in photosynthesis and transpiration as well as longer-term, acclimated variations. Key to this effort is explicitly separating the instantaneous responses of photosynthetic rates, and the slower acclimation of photosynthetic traits. The model also includes a dynamic optimization of stomatal conductance via the c_i:c_a ratio, which separates the rapid response to vapour pressure deficit (VPD) from a slower, acclimated response of the single parameter (ξ) of the stomatal optimality model to growth temperature.

A day-by-day diagnostic investigation has been carried out in order to optimize the behaviour of the resulting model at a half-hourly timestep. The model reproduces well the daily variations of GPP evaluated against FLUXNET observations, when forced with site-specific, half-hourly meteorological data from flux towers, and satellite data on the slowly varying fractional absorbed photosynthetically active radiation (fAPAR). Our approach accounts for the memory effect of past environmental conditions on photosynthetic traits, by introducing a daily average computation of temperature, solar radiation, VPD, CO₂ concentration and elevation. The results show that plants coordinate their biochemical capacities to match the maximum level of light during a day, optimizing to conditions near midday, when light is greatest. This is consistent with an interpretation of the co-limitation theory, whereby the V_cmax and J_max of leaves at any canopy level acclimate to the prevailing incident light. One implication is that the canopy is light-limited for most of the time. This is strongly supported by the diurnal time-course of observed GPP.