Fitness components versus total demographic effects: evaluating herbivore impacts on a perennial herb.

I examined the long-term effects of cryptic mollusk herbivory on seven fitness components in the perennial herb Lathyrus vernus and also calculated measures of total fitness effects. Natural correlations and experimental exclusion of mollusks showed that herbivory is associated with an increased probability of dying or staying dormant, poorer growth, and a lower probability of flowering. The average yearly reduction in population growth rate (lambda) caused by mollusks in the experiment was 0.14. The largest contribution to this decrease in total fitness occurred through a decreased survival of established plants. In contrast, seedling emergence and probability of flowering were the fitness components that were most affected in terms of relative change. The more important a life-cycle transition was for population growth rate in terms of its elasticity, the less it was affected by herbivore damage. These results suggest that simple analyses of the magnitudes of effects on individual components of plant performance are poor predictors of the magnitude of total fitness effects and tolerance to herbivory. This is because total fitness is differently sensitive to different phases of the life cycle and because plants strive to maintain the functions most important to fitness.