Insect floral herbivory can be a critical determinant of plant reproductive success in monocarpic species.  Optimal reproductive allocation in the presence of herbivory will be especially important in monocarpic plants, as a single flowering event determines their lifetime fitness. We are particularly interested in the effect of herbivory on the optimal resource allocation to the apical flower head versus to later flower heads that are often restricted by apical dominance.  Field experiments have demonstrated that apical heads that escape herbivory produced a large proportion of the plant’s seed set.  However, if the dominant apical head was destroyed by herbivores, and the apical dominance was broken, plants were unable to compensate for the lost seed production through seeds produced by later flower heads.  Therefore, investment in the apical head is risky.  We constructed a dynamic programming model to examine patterns of reproductive investment that maximize fitness in the native Platte thistle Cirsium canescens (Nutt.) and used field experiments to parameterize the model. This model allows us to examine conditions under which apical dominance (represented as initial disproportionate investment in a single developing flower head) results in a fitness advantage.

Results/Conclusions

Forward simulations using optimal resource decisions from the model suggest the timing of investment between the flower heads depends on the flower head survival probabilities.  When the probability of flower head survival is high, divergence in investment should occur early in the season, resulting in large differences in head sizes.  However, if survival probability is low plants should invest equally among flower heads at least two time periods later into the season, resulting in relatively smaller differences in flower head size.  This is interesting in the light of field data that suggest seed production in later flower heads does not compensate for loss of the apical head.  We are in the process of exploring how these model predictions change in the presence of multiple herbivores and with the incorporation of pollinator dynamics.  This model provides us with a deeper understanding of potential mechanisms underlying the flowering patterns we see in the field, and provides a useful framework for integrating species interactions into plant life history theory.