Environmental factors that govern the allocation of net primary production (NPP) between long-lived components (wood) and short-lived components (leaves, fine roots) are poorly understood yet essential when relating NPP to carbon stocks, especially among different plant functional types. We conducted a spatially synoptic analysis to investigate the relationships between NPP allocation and climate at the global scale. We ask a fundamental question in forest ecology and terrestrial carbon science: What environmental drivers influence NPP allocation?
Global forests (64.21 N–41.53 S).
Trees.
We investigated the relationships between field forest NPP data (n = 131) and 31 bioclimatic, meteorological, geographical, topographic and vegetation variables using machine learning.
These environmental variables accounted for 94%, 93% and 85% of the variation in canopy, woody and fine-root NPP fractions, respectively. Most of the important predictors were temperature-related. Allocation to the canopy was facilitated by warm and stable climates. In contrast, woody and fine-root carbon growth could endure cold and extreme temperatures.
Our analysis suggested that the most important drivers were annual mean (e.g. high canopy and fine-root NPP ratios in favourable climates), variation and extreme (e.g. significant woody and fine-root NPP ratios where light- and nutrient-limited) of temperatures. Overall, most carbon was stored in woody tissue and in a constant proportion to the canopy, which could be explained by allometric scaling and resource availability. Our results pointed to a ‘root-wood’ trade-off rather than the previously prevailing ‘root-leaf’ trade-off, but may vary regionally. For example, regarding particular plant functional types, ‘wood-leaf’ and ‘root-leaf’ were only evident in broadleaved and coniferous forests, respectively, to adapt to climates and compete for resources. Knowing the relationships between NPP allocation and the environment, we could assess forest carbon cycle dynamics in the face of climate change.