Herbivore Pressure on Goldenrods (Solidago Altissima): Its Variation and Cumulative Effects

Abstract

I used a combination of insect removal experiments and extensive censuses to estimate how the herbivore pressure on a native goldenrod, Solidago altissima, varied over a period of 6 yr. The pyrethroid insecticide, fenvalerate, was applied to experimental plots in an old field near Ithaca, New York, on staggered schedules so as to produce sets of plants that had experienced different levels of herbivory over different combinations of years. By comparing the performance of plants in the different treatments, I was able to measure both the contemporaneous and cumulative effects of insect herbivores on the growth and flowering of goldenrods. This protocol also provided the basis for estimating herbivore pressure (defined as the loss in plant performance that is associated with the presence of herbivores) by two independent procedures, the experimental suppression" approach and the correlation with herbivore load approach, which served as checks on various potential sources of error. I have taken this opportunity to illustrate the pitfalls and advantages of the two approaches within the context of a concrete example. The experiment demonstrated that goldenrods order their response to herbivory so as to favor maintenance of established individuals over seed production; both the proportion of stems that bloomed and the size of individual inflorescences were reduced during years when herbivore loads were too low to produce a measurable impact on the length and density of stems. Herbivore loads never reached levels sufficient to have detectable effects that carried over to influence plant performance in the following year. In addition, there was no evidence that insects were attracted to lush plants that had escaped herbivore attack the previous year. Relatively small loads, however, had effects that accumulated slowly over a period of years. Interestingly, ramets exhibited a different ordering in their response to the contemporaneous and cumulative effects of herbivory; the mass of inflorescences, the plant attribute that was most sensitive to contemporaneous herbivory, never responded to the effects of herbivory that accumulated over the 6 yr of the experiment. The experiment provided a means for inferring the impact of the herbivore loads that were obtained from censuses of 22 different goldenrod stands over the 6—yr period. These results show that goldenrod stands in central New York can be expected to experience relatively low herbivore loads (< 50 mg of herbivore dry mass per metre of stem) over 75% of the time. At these chronic levels herbivores cause a moderate reduction in the production of flowers but exert no measurable influence on stem growth. Infrequently (on the order of 5% of the time), herbivore loads can be expected to exceed 200 mg per metre of stem. At such loads herbivores can completely inhibit flowering and severely stunt vegetative growth. Such damaging loads are caused by population outbreaks of a few dominant insect species whose numbers usually wane before their cumulative impact threatens the goldenrod's persistence in the community. The herbivore pressures measured herein appear sufficient to act as potent agents for selecting between goldenrod genotypes.