Search

I present a meta‐analysis of plant responses to 48 nutrient addition experiments conducted with native species in naturally growing tropical forests, exclusive of mangrove forests. The added nutrients include nitrogen (N) in 36 experiments, phosphorus (P) in 33 experiments, calcium and potassium in one experiment each, and various mixtures of essential nutrients in the remaining experiments. I evaluate the hypotheses that nutrients limit tropical forest plants, nutrient limitation is stronger in successional than in old‐growth forests, P but not N is limiting in lowland forests, and N is limiting in montane forests. Responses to the most complete nutrient mix used in each experiment were strong for plant functions that contribute to aboveground production (Hedges' g averages 0.87) and nonsignificant for fine root biomass. Responses to N addition and to P addition were strong for tissue concentrations of the added element (Hedges' g averages 0.75 and 1.4, respectively), moderate for fine litter production (0.64 and 0.65, respectively), moderate to weak for plant growth (0.46 and 0.37, respectively) and nonsignificant for fine root biomass. Growth responses were stronger in successional than in old‐growth forests. All responses were unrelated to elevation. The 48 experiments included 30 factorial nitrogen‐phosphorus experiments that enable additional direct tests of the widely cited hypotheses that P limitation is stronger than N limitation in lowland forests and vice versa in montane forests. Both hypotheses were rejected. The N × P interaction effect was nonsignificant across the factorial experiments. In conclusion, nutrients clearly limit tropical forest plants. Limitation by N is widespread in both lowland and montane forests, and the same is true for P. Single experiments identify limitation by calcium and potassium, and correlative studies suggest limitation by calcium, potassium, and magnesium. The available evidence is consistent with the possibility that most macronutrients limit tropical forest plants; however, experiments focus almost exclusively on N and P. The way forward will include taking fuller advantage of existing nutrient addition experiments, siting new experiments strategically, and developing cost‐effective methods to assay responses to all of the essential nutrients soils supply to plants. [ABSTRACT FROM AUTHOR]