Age-size plasticity for reproduction in monocarpic plants.

Empirical and theoretical investigations of monocarpy have usually addressed the question of minimum or threshold sizes for reproduction. However, the range of flowering sizes observed in many monocarpic species is extraordinarily large (well beyond what can be called a "threshold"), and the sizes of flowering and nonflowering plants may overlap greatly. We attempt to explain these reproductive patterns in terms of optimal reaction norms predicted by simple deterministic life history models. We assume that individuals differ in their growth trajectories due to the heterogeneous quality of microsites and ask how the optimal age and size at flowering varies with environmental variation in growth and for different assumptions about fecundity and mortality. Under two very different growth functions (one with no age- or size-related decline in growth rate and another with such a decline as size approaches an asymptote), the optimal reaction norms imply considerable plasticity for size at reproduction, particularly when poor growth is associated with higher mortality or lower asymptotic size. Deterministic models such as these may be more applicable to long-lived than to short-lived monocarps, because fitness potential should be less affected by stochastic variability in yearly growing condition in the former than in the latter. We consider the case of a tropical monocarpic and masting tree species, Cerberiopsis candelabra (Apocynaceae), and show that our model results can account for wide ranges of reproductive size and overlap in size of flowering and nonflowering plants, in accord with observation. We suggest that empirical attention to norms of reaction across growth environments will be a more profitable approach than investigation of size thresholds per se.