Selective removal of insect herbivores from one plant species influences an old-field plant community

Background/Question/Methods

The question of whether insect herbivores influence plant communities was originally controversial; plants were expected to be limited primarily by competition rather than consumers. Several studies have now demonstrated that these herbivores can influence plant composition.  Most of these studies use “blanket removals”, excluding herbivores from the entire community. Resulting effects of herbivores have been assumed to arise primarily through changes in the competitive ability of dominant plant species, although interactions of herbivores with other community members might also be important. To determine if effects of herbivores on any particular plant species are important “selective removal” from individual plant species are required; experiments of this type are rare.  We selectively removed insect herbivores from a competitively inferior native plant (Solanum carolinense) in 20 experimental plots in North Florida old-field habitat, allowing natural herbivores in 20 control plots. These herbivore treatments were fully crossed with five initial planting densities of S. carolinense.  We removed herbivores by spraying all S. carolinense plants in each plot with insecticide. Treatments were maintained for five years. We measured plant community composition in the second and fifth year of the study.

Results/Conclusions

At planting (in 2007) S. carolinense was the only species in the plots but other species quickly established.  By 2008 mean S. carolinense cover was 15% and by the end of our 5-year experiment, S. carolinense was a minor component of the vegetation (mean cover = 3.3 %). Nevertheless, we found that removing herbivores from S. carolinense influenced the composition of the rest of the plant community, with species of intermediate abundance most strongly affected. Solanum carolinense density affected the plant community more strongly when herbivores were present.  The observed changes in community composition may have resulted from one of two potential mechanisms. First, herbivores may change S. carolinense demography or competitive ability, and S. carolinense may in turn influence other plants in the community through competition. Alternatively, removing herbivores from S. carolinense may reduce the number of generalist herbivores feeding on other members of the community; in this case our results would suggest that even a minor reduction in herbivore loads across multiple species can influence plant communities.