Estimating below-canopy light regimes using airborne laser scanning: An application to plant community analysis.

Light is a key driver of forest biodiversity and functioning. Light regimes beneath tree canopies are mainly driven by the solar angle, topography, and vegetation structure, whose three-dimensional complexity creates heterogeneous light conditions that are challenging to quantify, especially across large areas. Remotely sensed canopy structure data from airborne laser scanning (ALS) provide outstanding opportunities for advancement in this respect. We used ALS point clouds and a digital terrain model to produce hemispherical photographs from which we derived indices of nondirectional diffuse skylight and direct sunlight reaching the understory. We validated our approach by comparing the performance of these indices, as well as canopy closure (CCl) and canopy cover (CCo), for explaining the light conditions experienced by forest plant communities, as indicated by the Landolt indicator values for light ( ) from 43 vegetation surveys along an elevational gradient. We applied variation partitioning to analyze how the independent and joint statistical effects of light, macroclimate, and soil on the spatial variation in plant species composition (i.e., turnover, Simpson dissimilarity, ) depend on light approximation methodology. Diffuse light explained best, followed by direct light, CCl and CCo (  = .31, .23, .22, and .22, respectively). The combination of diffuse and direct light improved the model performance for compared with CCl and CCo (  = .30, .27 and .24, respectively). The independent effect of macroclimate on dropped from an of .15 to .10 when diffuse light and direct light were included. The ALS methods presented here outperform conventional approximations of below-canopy light conditions, which can now efficiently be quantified along entire horizontal and vertical forest gradients, even in topographically complex environments such as mountains. The effect of macroclimate on forest plant communities is prone to be overestimated if local light regimes and associated microclimates are not accurately accounted for.