An analytical model was used to describe the optimal nitrogen distribution. From this model, it was hypothesized that the non-uniformity of the nitrogen distribution increases with the canopy extinction rate for light and the total amount of free nitrogen in the canopy, and that it is independent of the slope of the relation between light saturated photosynthesis (Pm) and leaf nitrogen content (nL). These hypotheses were tested experimentally for plants with inherently different architectures and different photosynthetic modes. A garden experiment was carried out with a C3 monocot [rice, Oryza sativa (L.)], a C3 dicot [soybean, Glycine max (L.) Merr] a C4 monocot [sorghum, Sorghum bicolor (L.) Moensch] and a C4 dicot [amarantus, Amaranthus cruentus (L.)]. Leaf photosynthetic characteristics as well as light and nitrogen distribution in the canopies of dense stands of these species were measured. The dicot stands were found to have higher extinction coefficients for light than the monocot stands. Dicots also had more non-uniform N distribution patterns. The main difference between the C3 and C4 species was that the C4 species were found to have a greater slope value of the leaf-level Pm-nL relation. Patterns of N distribution were similar in stands of the C3 and C4 species. In general, these experimental results were in accordance with the model predictions, in that the pattern of nitrogen allocation in the canopy is mainly determined by the extinction coefficient for light and the total amount of free nitrogen.
Keywords: Canopy structure; Nitrogen allocation; Optimization; Photosynthesis; Photosynthetic pathway.