ALBERT

Notes: The three dimensional distribution of intercepted radiation, intercellular CO2 concentration (Ci) and late summer needle nitrogen (N) concentration were determined at the tips of all 54 branches in a 6·2-m-tall Pinus radiata D. Don tree growing in a New Zealand plantation. Measurements included above- and below-canopy irradiance, leaf stable carbon isotopic composition (δ13C) and tree canopy architecture. The radiation absorption component of the model, MAESTRO, was tested on site and then used to determine the branch tip distribution of intercepted radiation. We hypothesized that in branch tip needles: (i) the allocation of nitrogen and other nutrients would be closely associated with the distribution of intercepted radiation, reflecting carbon gain optimization theory, and (ii) Ci would predominantly reflect changes in photosynthetic rate (A) rather than stomatal conductance (gs), indicating that the increase in A for a given increase in N concentration was larger than the corresponding increase in gs. Needle nitrogen concentration was poorly related to intercepted radiation, regardless of the period over which the latter was calculated. At a given height, there was a large azimuthal variation in intercepted radiation but N concentration was remarkably uniform around the tree canopy. There was, however, a linear and positive correspondence between N concentration and δ13C and needle height above ground (r2 = 0·73 and 0·68, respectively). The very strong linear correspondence between N concentration and Ci (r2 = 0·71) was interpreted, using gas exchange measurements, as supporting our second hypothesis. Recognizing the strong apical control in P. radiata and possible effects of leaf nitrogen storage in an evergreen species, we propose that the tree leader must have constituted a very strong carbon sink throughout the growing season, and that the proximity of branch tip needles to the leader affected their photosynthetic capacity and nutrient concentration, independent of intercepted radiation. This implies an integrated internal determination of resource allocation within the tree and challenges the current convention that resources are optimally distributed according to the profile of intercepted radiation.