Biological Control Insect Use of Fertilized and Unfertilized Diffuse Knapweed in a Colorado Grassland

Theresponses and impacts ofÞve insect species that feedonCentaureadiffusaLamarck, diffuse knapweed, to soil nitrogen and phosphorus additions were studied in grasslands east of the ColoradoFrontRange.Wepredicted that fertilizationwasunlikely tohave adirect effect onherbivory but that if the insects preferentially select vigorous plants, increased tissue nitrogen, plant density, or plant mass resulting from fertilization would increase plant susceptibility to biological control insects. Fertilization caused a modest increase in the biomass of C. diffusa adults and the nutritive content of aboveground plant biomass. Larinus minutus Gyllenhal, a seed head weevil, was less abundant and exhibited reduced impacts on fertilized C. diffusa plants. Abundance of gall ßy larvae, Urophora spp., was negatively inßuenced by the presence of L. minutus and had no detectable effect on seed abundance. No positive or negative impacts of root feeders on knapweed were observed, but the root-feeding Cyphocleonus achates (Fahraeus) selected plants in fertilized plots, whereas the root phytophage Sphenoptera jugoslavica Obenberger was more common on knapweed in unfertilized plots. Our results indicate that these specialist biological control insects exhibit little preference based on plant biomass, density, or nutritive quality. However, variables that affect the timing of knapweed seed production, such as fertilization, mowing, and grazing, reduce insect impacts and may explain some of the regional variation in effectiveness of these insects as biological controls on knapweed.