Tropical forests are faced with several shifting climate factors because of global warming.  Light intensity and the distribution of direct - diffuse light are two key factors expected to change as many tropical forest experience drying, decreased cloud cover, and increased dry season lengths.  Understanding the direct and interactive effect of changes in light regime on forest function, composition, biomass, and size structure are key questions in tropical forest ecology.  Shifting light environments will first alter patterns of photosynthetic production.  This photosynthetic effect will, in turn, alter demography—growth and mortality—of the plant community, potentially differently for different forest strata (size-classes).  A great challenge, though, is that addressing the (i) direct photosynthetic and (ii) size-structured demographic effects of light limitation is extremely difficult without detailed information about light and leaf area variation in three dimensions over large tracts of forest, as well as the locations and spatial extents of individual trees.  Here we ask if combining high resolution aerial LiDAR point-scatter data and a similar higher resolution (but lower extent) ground-based LiDAR dataset with long-term tree plot surveys, that include crown maps, can effectively address the photosynthetic and demographic role of light limitation in the central Amazon.

Results/Conclusions

We find that the intersection of high-resolution LiDAR data and plot studies can effectively relate light limitation to ecosystem photosynthesis and size structured tree demography.  However, a few sources of uncertainty must be overcome to better bring the weight of this potentially exceptional approach to bear on the question of climate change in tropical forests.  Key sources of uncertainty include (i) the implementation and testing of light transmission models, (ii) accurately geo-referencing and locating individual trees in LiDAR datasets, and (iii) attributing quanta of leaf area and leaf photosynthetic functional traits to individual trees.  Furthermore, we find that the ability of LiDAR data to evaluate ensemble characteristics such as the vertical distribution of vegetation and light availability (and the variability of these distributions) is relatively unaffected by these sources of uncertainty; these ensemble characteristics can be combined with size structured ecosystem models to drastically improve regional estimates of forest processes.