RESUMO
Nutrient resorption is an important mechanism for nutrient conservation and can maintain ecosystem stoichiometry. Here, we examined the global-scale variation of nitrogen resorption efficiency (NRE) and phosphorus resorption efficiency (PRE) by analyzing observations from 218 published papers. We used Pagel's λ to test the phylogenetic limitation on NRE and PRE and applied the random forest model to assess biotic and abiotic drivers, which included climate, soil, species characteristics, and topographical factors, and predicted the global NRE and PRE distributions. We found that NRE and PRE had oppositing trends among climatic zones, plant functional groups, and foliar nitrogen (N) to phosphorus (P) ratios. Nutrient resorption was higher in ectomycorrhizal trees than in arbuscular mycorrhizal trees. Moreover, foliar NRE and PRE were not linked to phylogeny. On average, the random forest overall explained 38 % (21 %-55 %) variation in NRE and 36 % (16 %-55 %) variation in PRE. Both NRE and PRE varied greatly with climate and soil organic carbon (SOC). The spatial variation of NRE and PRE was coupled to N-limitation and P-limitation, respectively. Our evaluation of the factors that influenced NRE and PRE and their global distributions, and our novel approach for evaluating plant utilization of nutrients, advances our understanding of the relative stability of ecosystem randomness in forest ecosystems and the global forest nutrient cycle.
