RESUMO
The growth differentiation balance hypothesis (GDBH) provides a framework that predicts a trade-off between costs of secondary metabolites (SMs) relative to the demand for photosynthate by growth. However, this hypothesis was developed using empirical evidence from plant species in northern boreal and temperate systems, leaving its applicability to species under different abiotic and biotic conditions questionable and generalizations problematic. The objective of this study was to investigate whether the GDBH explains allocation to SMs in the deciduous African savanna woody species C. apiculatum along a 6-point N gradient. The cornerstone prediction of the GDBH, i.e., the parabolic response in SMs along the N gradient, was not observed, with secondary metabolism showing compound-specific responses. Quercetin, myricetin, and kaempferol glycoside concentrations, all produced via the same pathway, responded differently across the N gradient. Flavonol glycoside, cinnamic acid, and quercetin glycoside concentrations decreased as N increased, which provides partial support for the carbon nutrient balance hypothesis. Simulated herbivory had no effect on photosynthesis, decreased foliar N and consequently increased C:N ratio, but did not induce an increase in SMs, with condensed tannins and flavonol glycosides being unaffected. Defoliated plants at low N concentration compensated for lost biomass, which suggests a tolerance response, but as predicted by the limiting resource model, plants at higher N concentration were evidently C limited and thus unable to compensate. Our results show that the GDBH does not explain allocation to SMs in C. apiculatum, and suggest that mechanistic explanations of plant allocation should consider the integrative defensive effect of changed SMs.