Effects of simulated herbivory in three old field Compositae with different inflorescence architectures.

The effects of simulated herbivory (early or late defoliation and cutting of the flowering shoot) on the growth and reproduction of three species of monocarpic composite forbs (Crepis pulchra, Picris hieracioides and C. foetida) with different inflorescence architectures were studied in experimental plots. For the three species studied, early defoliation had no significant effect on subsequent growth. In contrast, late defoliation, occurring at the start of the season of drought, had a negative effect on growth and reproduction in the two Crepis species, particularly C. foetida, but had less effect on P. hieracioides. Sexual biomass was more clearly affected by late defoliation than was vegetative biomass, although the effects differed markedly among species possibly as a result of differences in phenology. Clipping the flowering shoot removed about 3 times less biomass than late defoliation and had little effect on vegetative biomass. It had much greater effects on the sexual biomass in P. hieracioides and C. pulchra, and resulted in the production of many shoots sprouting from the rosette, allowing the treated plants to regain a vegetative biomass close to that of control plants. Clipping did however lead to the production of shorter shoots and a reduction in the number of capitula formed. In C. foetida, much branching occurred even when the main shoot was not cut; the architecture of individual plants was therefore only slightly changed by clipping the apical bud and the sexual biomass of this species was not affected by ablation of the flowering shoot. Overcompensation was found in only two families of C. pulchra for vegetative biomass. No over-compensation was found for sexual biomass, despite an increase in the number of flowering shoots in C. pulchra and P. hieracioides following clipping. However situations close to compensation for the vegetative biomass in the three species and in P. hieracioides for the sexual biomass were recorded. The response of the three study species to simulated herbivory were related to their architecture and to the time of defoliation.