Leaf boundary layer conductance of two native species in southeast Spain

Leaf boundary layer conductance was measured in the field using heated leaf replicas placed in the canopies of two of the three main semi-arid species found in the Rambla Honda (Almeria, South East of Spain), Retama sphaerocarpa (L.) Boiss, and Stipa tenacissima L. R. sphaerocarpa is a woody perennial bush with approximately cylindrical cladodes arranged on randomly orientated stems. In contrast S. tenacissima is a perennial tussock grass with leaves that can roll during periods of water stress. Cylindrical leaf replicas were constructed with dimensions similar to the leaves of S. tenacissima and cladodes of R. sphaerocarpa. The relationship between wind speed (u) measured outside the plant canopy and boundary layer conductance (g(a)) was analysed on the basis of an equation, derived from engineering formulae describing forced convection for a cylinder in laminar flow. The coefficients a and b were derived empirically from experimental data for replicas placed within and outside the canopies of the two species. Measured boundary layer conductance in the open tended to be higher than that estimated from engineering equations resulting in higher estimates of a. There was no significant difference in the estimate of b between replicas in the open and the engineering value. In the R. sphaerocarpa canopy the estimate of a tended to decrease with depth into the canopy but there was no significant difference in estimates of b between replicas inside and outside the canopy. Inside the S. tenacissima canopy both coefficients a and b tended to be significantly lower than those derived from engineering equations. Significant differences between the parameters of the relationship inside and outside the canopy showed the difficulty in deriving a unique equation to calculate the boundary layer conductance of a whole plant from wind speed measured outside the canopy. This work confirms that serious errors can be made if engineering equations are applied without a better knowledge of the relationship between boundary layer conductance and wind speed. The results suggest that reduction of average wind speed by shelter within the canopy and increased turbulence are important in influencing leaf boundary layer. The heated leaf replica method used in this work provides a simple and effective system for measuring leaf boundary layer conductance for plants in the field, and is preferable to the use of uncalibrated empirical equations. These results can be related to the importance of leaf shape and canopy structure in determining water use of natural vegetation in semi-arid regions. R. sphaerocarpa has an open canopy allowing a better penetration of air flow. S. tenacissima has a dense canopy producing shelter decreasing wind speed rather than increasing turbulence. While both species have similar shaped transpiring plant parts, the strategy of S. tenacissima to maintain dead materials as part of the plant canopy could be important in control of water loss.