Boundary layer conductance for contrasting leaf shapes in a deciduous broadleaved forest canopy

A new method of constructing light, flexible and more realistic replica leaves for continuous determination of leaf boundary layer conductance to heat transfer (gbh) was developed and tested in a mature oak (Quercus robur L.) and sycamore (Acer pseudoplatanus   L.) tree canopy. The replicas were used to determine the difference between oak and sycamore leaf gbh in exposed sites in the upper canopy, the relationship of gbh with wind speed, the seasonal changes in gbh in the canopy as leaf cover developed, and values for Ω  , the decoupling coefficient. The replicas showed similar gradients in temperature at their margins to those in real leaves. When exposed, the gbh of the larger sycamore leaves was 66% of that of oak leaves under the same conditions. Linear relationships were found with gbh and wind speed across the measured range of 0.3–3.5 m s−1, and flow in the replica boundary layers was laminar in all conditions. The leafless canopy produced a substantial sheltering effect, reducing gbh by 12–28% in light winds. Sycamore replicas in the leafed canopy showed a 19–29% lower gbh at a given external wind speed than when outside, but there was little difference between ‘sun’ and ‘shade’ position shoots, because of the density of the shoots, and the close branching pattern. In contrast, in oak gbh at a given wind speed was 15–21% lower for ‘sun’ leaves than that for replicas outside the canopy, with a larger reduction (approximately 28%) in denser ‘shade’ sites. Although wind speed in the canopy was often low, leaves of both species were usually well coupled to the canopy airstream (Ω < 0.3). Sun leaves were substantially less well coupled than shade leaves, despite the lower shelter effect, because of their higher stomatal conductance values. In the lightest winds (<0.5 m s−1) and with high stomatal conductance, coupling may on many occasions be poor for sun leaves, particularly for the larger sycamore leaves.