Global‐scale climatic controls of leaf dry mass per area, density, and thickness in trees and shrubs

Leaf dry mass per unit area (LMA) is a product of leaf thickness (T) and of density (D). Greater T is associated with greater foliar photosynthetic rates per unit area because of accumulation of photosynthetic compounds; greater D results in decreased foliage photosynthetic potentials per unit dry mass because of lower concentrations of assimilative leaf compounds and decreases in intercellular transfer conductance to CO₂. To understand the considerable variation in T and D at the global scale, literature data were analyzed for 558 broad‐leaved and 39 needle‐leaved shrubs and trees from 182 geographical locations distributed over all major earth biomes with woody vegetation. Site climatic data were interpolated from long‐term world climatologies (monthly precipitation, surface temperature) or modeled using the Canadian Climate Center Model (monthly global solar radiation). Influences of total annual precipitation (WT), precipitation of the driest month (Wₘᵢₙ), monthly mean precipitation of the three driest months in the year (W₃ₘᵢₙ), highest monthly precipitation (Wₘₐₓ), precipitation index ([Wₘₐₓ − Wₘᵢₙ]/WT), mean, minimum, and maximum annual monthly temperatures, and daily annual mean global solar radiation (R) on LMA, D, and T were tested by simple and multiple linear and log‐linear regression analyses. In broad‐leaved species, LMA and T increased with increasing R and mean temperature and scaled weakly and negatively with precipitation variables, but D was negatively related only to precipitation. Similar relationships were also detected in needle‐leaved species, except that, in multiple regression analysis, precipitation did not significantly influence leaf thickness, and R was positively related to D. Given that increases in LMA and T are compatible with enhanced photosynthetic capacities per unit leaf area, but also with greater costs for construction of unit surface area, positive effects of solar irradiance and surface temperature on these variables are indicative of shorter leaf pay‐back times in conditions of higher irradiance and temperature allowing construction of leaves with higher photosynthetic potential. To gain insight into the scaling of leaf density with site aridity, correlations of D with the leaf elastic modulus close to full turgor (ε) and with the leaf osmotic potentials (π) at full and zero turgor were analyzed. Both low π, which is compatible with low leaf water potential, and high ε, which permits large adjustment of leaf water potential with small changes in leaf water content, may facilitate water uptake from drying soil. Leaf elastic modulus was independent of T and was weakly related to LMA; but there were close positive associations of ε with D and leaf dry to fresh mass ratio, which is an estimate of apoplastic leaf fraction. Consequently, changes in D bring about modifications in leaf elasticity and allow tolerance of water limitations. Across all the data, ε and the estimates of π were negatively related. However, given that π varied only fourfold, but ε 10‐fold, I conclude that osmotic adjustment of leaf water relations is inherently limited, and that elastic adjustment resulting from changes in leaf structure may be a more important and general way for plants to adapt to water‐limited environments.