Leaf N:P ratio does not predict productivity trends across natural terrestrial ecosystems

Nitrogen (N) and phosphorus (P) are crucial nutrients for regulating plant growth. The classic growth rate hypothesis (GRH) proposes that fast-growing organisms have lower N:P ratios, and it is promising to predict net primary productivity (NPP) using the leaf N:P ratio at the community level (N:P_Com). However, whether leaf N:P ratio can predict NPP in natural ecosystems on a large scale remains nebulous. Here, we systematically calculated leaf N:P_Com (community biomass-weighted mean and species arithmetic mean) using the consistently measured data of 2192 plant species-site combinations and productivity (biomass-based aboveground NPP and flux-based NPP) in 66 natural ecosystems in China. Unexpectedly, leaf N:P_Com hardly predicted productivity in natural ecosystems due to their weak correlation, although significantly negative or positive relationships across different ecosystems were observed. The ambiguous relationship between leaf N:P and species dominance reflected a luxury consumption of N and P in turnover and structure in natural communities, unlike what GRH suggests. Climate, soil, and leaf nutrients (rather than N:P) influenced productivity, which highlighted the importance of external environment and nutrient constrains. Our findings pose a major challenge for leaf N:P_Com as a direct parameter in productivity models and further question the direct application of classic hypotheses in short-term experiments or model species to long-term and complex natural ecosystems.