How should leaf area, sapwood area and stomatal conductance vary with tree height to maximize growth?

Conventional wisdom holds that the ratio of leaf area to sapwood area (L/S) should decline during height (H) growth to maintain hydraulic homeostasis and prevent stomatal conductance (g(s)) from declining. We contend that L/S should increase with H based on a numerical simulation, a mathematical analysis and a conceptual argument: (1) numerical simulation--a tree growth model, DESPOT (Deducing Emergent Structure and Physiology Of Trees), in which carbon (C) allocation is regulated to maximize C gain, predicts L/S should increase during most of H growth; (2) mathematical analysis--the formal criterion for optimal C allocation, applied to a simplified analytical model of whole tree carbon-water balance, predicts L/S should increase with H if leaf-level gas exchange parameters including g(s) are conserved; and (3) conceptual argument--photosynthesis is limited by several substitutable resources (chiefly nitrogen (N), water and light) and H growth increases the C cost of water transport but not necessarily of N and light capture, so if the goal is to maximize C gain or growth, allocation should shift in favor of increasing photosynthetic capacity and irradiance, rather than sustaining g(s). Although many data are consistent with the prediction that L/S should decline with H, many others are not, and we discuss possible reasons for these discrepancies.