Resource availability at the rosette stage and apical dominance in the strictly biennial Erysimum strictum (Brassicaceae)

In biennial plants, the age of flowering is constrained, but size at flowering is highly variable. This suggests that performance at the flowering stage depends largely on growth conditions at the rosette stage. We examined this possibility using Erysimum strictum P. Gaertn., B. Mey., and Scherb. (Brassicaceae), a strictly biennial herb, the reproductive output of which increases with increasing plant height and branch number. In a common garden experiment, we defoliated (50% of leaves removed twice) and fertilized (three times) individual plants at the rosette stage and studied their performance at the flowering stage in the following year. Rosette defoliation adversely affected all performance measures except seed number per fruit and seed weight. Fertilization did not alleviate these effects. Defoliation reduced seed set by 48% in fertilized plants and 29% in unfertilized plants. Fertilization stimulated branch production from the rosette base but did not significantly affect plant height. These observations suggest that, in the case of basally unbranched plants, apical dominance by the leading stalk suppresses the axillary meristems at the rosette base. Fertilization at the rosette stage can break this suppression. The induction for breakage presumably occurs before bolting since, in our earlier experiments, neither fertilization nor apical damage at the flowering stage stimulated branching from the base. Erysimum strictum is likely to be selected for fast vertical growth at the start of bolting, and hence plant height is a less plastic trait with respect to resource availability than branch number. Regression analysis suggested that, in response to rosette fertilization, small plants invest in height growth instead of branching, whereas large plants to a greater extent invest their supplemental resources in vigorous branching. Consequently, resource availability at the rosette stage influences apical dominance at the flowering stage. Unexpectedly, however, improved resource availability did not alleviate the cost of simulated rosette-stage herbivory.