3D upscaling of transpiration from leaf to tree using ground-based LiDAR: Application on a Mediterranean Holm oak (Quercus ilex L.) tree

The reliability of plant upscaling methods strongly depends on a precise calculation of the intercepted radiation, which in turn depends on the three-dimensional distribution of the plant biomass. A three-dimensional LiDAR upscaling procedure (LUP) based on a fine description of plant structure is proposed. The voxel-based vegetative elements distribution of a Quercus ilex L. tree was acquired with a ground-based LiDAR system. These preprocessed LiDAR data were imported into a ray tracing program in order to simulate the light environment through the crown during 15 days on a half hour time step. The obtained light environments were then used as input for a Jarvis-type conductance model in order to calculate the transpiration per voxel by inverting the Penman-Monteith equation. The approach reproduced the vertical LAI profile of the tree and the exponential extinction of light through the canopy. It also provided the possibility to observe the contribution of any voxel inside the crown to the total transpiration. The LAI of the tree measured by the LiDAR system resulted in a value of 3.97, comparable with LAI measured using the hemispherical photography and a LICOR LAI 2000,3.64 and 4.1 respectively. Maximum conductance (g(s.max)) varied between 70 and 210 mmol H(2)O m(-2) s(-1) depending on the average daily intercepted radiation. The total transpiration rate of the tree was obtained by integrating transpiration of all voxels and validated by comparison with direct measurements of sap flow in August, September and October with a different water availability, 10%, 15% and 19%. The method yielded an R 2 of 0.90, without the need to parameterize coefficients with the direct measurements. It is shown how the most shaded parts of the canopy (voxels that intercepted 0-20% of the total daily radiation) contributed the most to the total transpiration because of the wider surface of this class. The model also predicts a ratio between leaf maximum conductance (g(s)) of sun leaves and canopy maximum conductance (G(smax)) of 2.56. Both these results suggest that shade leaves instead of sun leaves should be chosen as sample leaves for upscaling purposes. The reliability of the developed method proved to be independent of the horizontal heterogeneity of the canopy. This independence is maid possible by the choice to describe the 3D light environment with small voxels (0.1 m(3)), hence this methods intrinsically takes into account gaps and leaf clumping. Given the heterogeneity of the canopy, 10-26% of the initial incoming radiation was allowed to reach the understorey directly. The combination of structural LiDAR data with ecophysiological measurements proved to be a valid tool for upscaling. However some steps in the procedure could be optimized, leaving space for further improvements.