Control of transpiration from the upper canopy of a tropical forest: the role of stomatal, boundary layer and hydraulic architecture components

ABSTRACT Concurrent, independent measurements of stomatal conductance ( g s ), transpiration ( E ) and microenvironmental variables were used to characterize control of crown transpiration in four tree species growing in a moist, lowland tropical forest. Access to the upper forest canopy was provided by a construction crane equipped with a gondola. Estimates of boundary layer conductance ( g b ) obtained with two independent methods permitted control of E to be partitioned quantitatively between g s and g b using a dimensionless decoupling coefficient (Ω) ranging from zero to 1. A combination of high g s ( c. 300–600 mmol m −2 s −1 ) and low wind speed, and therefore relatively low g b ( c. 100–800 mmol m −2 s −1 ), strongly decoupled E from control by stomata in all four species (Ω= 0.7–0.9). Photosynthetic water‐use efficiency was predicted to increase rather than decrease with increasing g s because g b was relatively low and internal conductance to CO 2 transfer was relatively high. Responses of g s to humidity were apparent only when the leaf surface, and not the bulk air, was used as the reference point for determination of external vapour pressure. However, independent measurements of crown conductance ( g c ), a total vapour phase conductance that included stomatal and boundary layer components, revealed a clear decline in g c with increasing leaf‐to‐bulk air vapour pressure difference ( V a because the external reference points for determination of g c and V a were compatible. The relationships between g c and V c and between g s and V s appeared to be distinct for each species. However, when g s and g c were normalized by the branch‐specific ratio of leaf area to sapwood area (LA/SA), a morphological index of potential transpirational demand relative to water transport capacity, a common relationship between conductance and evaporative demand for all four species emerged. Taken together, these results implied that, at a given combination of LA/SA and evaporative demand scaled to the appropriate reference point, the vapour phase conductance and therefore transpiration rates on a leaf area basis were identical in all four contrasting species studied.