Abstract
AbstractThe large grove-forming tropical treeMicroberlinia bisulcata(Fabaceae) at Korup, Cameroon, shows strong mast fruiting. Reproductive allocation is considerable. The site has very nutrient-poor soil. To test the nutrient resource limitation hypothesis, phenological recordings between 1989 and 2017 were matched with climate variables, and analyzed using logistic time-series regression. Masting happened mostly on 2- or 3-year cycles. A strong predictor was mean daily rainfall in the dry season: low in the current year of masting and high in the year prior. Less strongly predictive was the increase in dry season radiation between prior and mast years. Masting events showed no relationship to annual stem increment, nor with local plantation yields. Later, the normally heavy mastings became moderate after two attacks by caterpillars. Collated studies of fallen leaf nutrient concentrations showed that P increased markedly, K rose and fell, but N and Mg changed little, in the inter-mast interval. P and K were likely being accumulated and stored, and then triggered masting events when internal thresholds were crossed. The drier season prior to masting enabled a rise in C, and the wetter season the year before, with higher soil moisture, enabled better acquisition and uptake of nutrients by roots and mycorrhizas. The main storage of P may be in bark and branches, that for K on soil organic-colloids. A rooting-fruiting trade-off in C allocated over a minimal 2-year cycle is implied. Hypothesized is that synchrony among masting trees may be achieved, in part, by an equilibration of P across the mycorrhizal network and possibly root grafts. The long-term driver appears to be the inherent year-to-year stochasticity of dry-season rainfall, realization of which leads to an important refinement of the hypothesis. Life history strategy linked to nutrient resource dynamics provides a plausible explanation and more advanced hypothesis for the masting events observed.
Publisher
Cold Spring Harbor Laboratory