Unmanned aerial vehicles (UAVs) have opened new opportunities for measuring 3D forest canopy structure from aerial imagery. Image data collected with a UAV can be processed to generate detailed information on local canopy structure around an individual tree, which may be a useful proxy for the amount of competition that tree experiences from its neighbours. Structural indices of competition traditionally have been derived from ground-based plot data, and it is not clear whether aerial canopy data are as effective as ground data for modelling the effects of competition on individual-tree growth. To assess the effectiveness of aerial competition indices, we compare the relative performance of four ground-based competition indices derived from plot data, five canopy-based competition indices derived from UAV data, and one hybrid index that uses both data types, for predicting the radial growth of two northern tree species (white spruce, Picea glauca; lodgepole pine, Pinus contorta).
Results/Conclusions
Ground-based and canopy-based competition indices were both represented among the top-performing models for each species, but no single index was unambiguously favoured over all others. Of the ten competition indices we considered, the mean canopy height within 15 m of a subject tree had the strongest performance across both species, including the best performance for lodgepole pine (R2=0.29) and third-best performance for white spruce (R2=0.42). Models with mean canopy height also revealed interactions between competition and soil moisture, with growth reductions from competition limited to dry sites for white spruce and to mesic sites for lodgepole pine. Overall, our comparison showed that canopy-based metrics such as mean canopy height can be at least as effective as traditional ground-based metrics for measuring the effects of local competition on tree growth. As a new research tool in forest ecology, UAVs thus offer a valuable approach for measuring neighbourhood crowding and its effects on the performance of individual trees.