Global climatological determinants of canopy height

Qiuyan Yu, Rolston St. Hilaire, Wenjie Ji, Julius Anchang and Niall P. Hanan, Plant and Environmental Sciences, New Mexico State University, Las Cruces, NM, Lara Prihodko, Animal and Range Sciences, New Mexico State University, Las Cruces, NM, Michael G. Ryan, Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO

Background/Question/Methods:
Canopy height is a critical determinant of ecosystem function and net primary production, biomass and terrestrial carbon stocks. Improved understanding of the role of climate, soils, physiology and phylogeny in determining maximum tree height will provide critical inputs to carbon and dynamic vegetation models, and will enrich sustainable ecosystem management. Site-scale observations show clear effects of water availability on maximum canopy height. However, at macro-ecological scales the hydrologic controls on maximum tree height are neither well characterized nor well understood. This study leverages Global Ecosystem Dynamics Investigation (GEDI) data to characterize worldwide patterns in canopy height and assess how these patterns are related to mean annual precipitation (MAP) and rainfall seasonality. We collected GEDI canopy height estimation within gridded plots (20km × 20km) at all the intersections of latitudes and longitudes in the coverage of GEDI (51.6°S – 51.6°N). An enveloping approach was adopted for analysis and interpretation of the massive GEDI datasets using quantile (90th percentile) regression to empirically define the relationship between potential canopy heights, MAP and rainfall seasonality. While maximum canopy height can be affected by local topo-edaphic condition and disturbances, with very large canopy height datasets, the enveloping function holds the potential to capture the relationship between potential canopy height and rainfall at regional and global scales.
Results/Conclusions:
MAP is a primary determinant of potential canopy height worldwide. As expected, potential tree heights increase for the low range of MAP, and saturates at higher MAP. However, we also detect distinct step-functions associated with vegetation types, indicating a role for phylogenetic controls. Rainfall seasonality is generally less significant in regulating potential canopy height, with stronger impacts in drier regions. Our analyses showed interpretable relationships between rainfall and canopy height using a simple and robust approach and high-resolution global-scale canopy height data. This knowledge has the potential to enrich dynamic vegetation models, and constrain predictions of changes in ecosystem structure in relation to climate change.