Regional and historical factors supplement current climate in shaping global forest canopy height

Abstract

Canopy height is a key factor that affects carbon storage, vegetation productivity and biodiversity in forests, as well as an indicator of key processes such as biomass allocation. However, global variation in forest canopy height and its determinants are poorly known. We used global data on Light Detection and Ranging-derived maximum forest canopy height (Hmax) to test hypotheses relating Hmax to current climate (water availability, ambient energy and water-energy dynamics), regional evolutionary and biogeographic history, historical climate change, and human disturbance. We derived Hmax for 32 304 forested 55-km grid cells using 1-km global canopy height data (maximum height of 1-km cells within a 55-km grid). Variation in Hmax was related to latitude and biomes, along with environmental and historical variables. Both spatial and non-spatial linear models were used to assess the relative importance of the different hypothesized factors. Hmax was inversely related to latitude (i.e. tall canopies at the equator), but with high geographical variability. Actual evapotranspiration and annual precipitation were the factors most correlated to Hmax globally, thus supporting the water-energy dynamics hypothesis. However, water limitation emerged as a key factor in tropical and temperate biomes within specific geographic regions, while energy limitation was a more important factor in boreal regions where temperature is more limiting to trees than water. Hmax exhibited strong variation among biogeographic regions, supporting the role of regional evolutionary and biogeographic history in structuring broad-scale patterns in canopy height. Furthermore, there were divergent relationships between climate and Hmax between the Southern and Northern Hemispheres, consistent with historical evolutionary contingencies modulating these relationships. Historical climate change was also related to Hmax, albeit not as strongly, with shorter canopy heights where late-Quaternary climate has been less stable. In contrast, human disturbance was only weakly related to Hmax at the scale (55 km) examined here. Synthesis. This study confirms that forest canopy height is strongly controlled by current climate, but also provides evidence for an important supplementary role for regional-historical factors. This highlights the importance of considering evolutionary and biogeographic history for achieving a comprehensive understanding of forest ecosystem properties. This study confirms that forest canopy height is strongly controlled by current climate, but also provides evidence for an important supplementary role for regional-historical factors. This highlights the importance of considering evolutionary and biogeographic history for achieving a comprehensive understanding of forest ecosystem properties.