Optimality and nitrogen allocation in a tree canopy.

Physical and functional properties of foliage were measured at a variety of microsites in a broad-leaved Nothofagus fusca (Hook. f.) Ørst. canopy. The light climate of the foliage at these sites was monitored for 39 days in the late spring and early summer with in situ sensors. Foliage nitrogen content (N), mean leaf angle, and gas exchange characteristics were all correlated with the amount of light reaching the microsites during foliage development. Foliage N content on a leaf area basis ranged between ~1 and 2.5 g N m(-2) and was highest at the brightest sites. Light-saturated photosynthetic rates ranged between ~4 and 9 micro mol m(-2) s(-1), increasing from the darkest to brightest sites. A biochemical model of photosynthesis was fitted to foliage characteristics at the different microsites and used to integrate foliage assimilation among the sites over 39 days. The actual arrangement of foliage physiological characteristics in the observed microsites led to higher total canopy rates of net assimilation than > 99% of the combinations of observed foliage characteristics randomly assigned to the observed microsites. Additional simulations first related the maximum rates of electron transport (J(max)), ribulose bisphosphate turnover (V(c,max)), and dark respiration (R(d)) of Nothofagus fusca foliage to nitrogen content and then allowed foliage N (and consequently leaf gas exchange characteristics) to vary across the canopy. The observed N allocation pattern results in greater total canopy assimilation than uniform or > 99% of the simulations with random distributions of N among the microsites (constrained so that the total N allocated was equivalent to that observed in the microsites). However, the observed pattern of N allocation places less N in the brightest microsites and results in substantially less total assimilation than a simulated canopy in which N was allocated in an optimal manner where the N distribution is such that the partial derivative of leaf assimilation (A) with respect to leaf nitrogen content, partial differential A/ partial differential N, is constant among microsites. These results suggest that other factors such as wind or herbivory reduce the integrated assimilation of high-N foliage relatively more than lower-N foliage and that a partial differential A/ partial differential N optimality criteria based only on formulations of leaf gas exchange overestimate canopy assimilation.