In big-leaf models of canopy photosynthesis, the Rubisco activity per unit ground area is taken as the sum of activities per unit leaf area within the canopy, and electron transport capacity is similarly summed. Such models overestimate rates of photosynthesis and require empirical curvature factors in the response to irradiance. We show that, with any distribution of leaf nitrogen within the canopy (including optimal), the required curvature factors are not constant but vary with canopy leaf area index and leaf nitrogen content. We further show that the underlying reason is the difference between the time-averaged and instantaneous distributions of absorbed irradiance, caused by penetration of sunflecks and the range of leaf angles in canopies.
These errors are avoided in models that treat the canopy in terms of a number of layers – the multi-layer models. We present an alternative to the multi-layer model: by separately integrating the sunlit and shaded leaf fractions of the canopy, a single layered sun/shade model is obtained, which is as accurate and simpler. The model is a scaled version of a leaf model as distinct from an integrative approach.