Abstract
BACKGROUND AND AIMS: The nitrogen (N) to phosphorus (P) ratio (N:P) has been widely used as a threshold for identifying nutrient limitations in terrestrial plants; however, the associated reliability has not been well assessed. METHODS: The uncertainty of nutrient limitations caused by the N:P threshold was evaluated using two approaches: fertilization experiments synthesized across multiple ecosystems; and random sampling simulation of the impacts of different nutrient sufficiencies and deficiencies. KEY RESULTS: The fertilization experiment data indicated that the types of nutrient limitation determined via N:P thresholds were partly inconsistent with the growth responses observed under N and P additions, i.e. under N:P thresholds of 14 and 16 (or 10 and 20), 32.5 % (or 16.2 %) of the data were inconsistent between these two. The random sampling simulation suggested that N:P thresholds may indicate N (or P) limitations when leaf N (or P) content is sufficient, whereas these thresholds may not indicate N (or P) limitations when leaf N (or P) content is deficient. The error risks calculated from the sampling simulation presented large fluctuations at small sample sizes and decreased as the thresholds of nutrient content sufficiency (or deficiency) increased (or decreased). The N:P thresholds of 10 and 20 showed lower error risks than the thresholds of 14 and 16. CONCLUSIONS: These findings highlight that canonical N:P thresholds have the potential to introduce a large uncertainty when used to detect plant nutrient limitations, suggesting that the error risks should be cautioned in future studies.