Non-target effects of an introduced biological control agent on deer mouse ecology

Release of exotic insects as biological control agents is a common approach to controlling exotic plants. Though controversy has ensued regarding the deleterious direct effects of biological control agents to non-target species, few have examined the indirect effects of a ”well-behaved” biological control agent on native fauna. We studied a grassland in west-central Montana infested with spotted knapweed (Centaurea maculosa) to examine the effects of knapweed invasion and two gall flybiological control agents (Urophora affinis and U. quadrifasciata) on the native deer mouse (Peromyscus maniculatus). Stomach-content analysis revealed that Urophora were the primary food item in Peromyscus diets for most of the year and made up 84–86% of the winter diet. Stomach contents indicated that wild-caught mice consumed on average up to 247 Urophora larvae mouse^–1 day^–1, while feeding trials revealed that deer mice could depredate nearly 5 times as many larvae under laboratory conditions. In feeding trials, deer mice selected knapweed seedheads with greater numbers of galls while avoiding uninfested seedheads. When Urophora larvae were present in knapweed seedheads, deer mice selected microhabitats with moderately high (31–45% cover) and high knapweed infestation (≥46% cover). After Urophora emerged and larvae were unavailable to Peromyscus, mice reversed habitat selection to favor sites dominated by native-prairie with low knapweed infestation (0–15%). Establishment of the biological control agent, Urophora spp., has altered deer mouse diets and habitat selection by effecting changes in foraging strategies. Deer mice and other predators may reduce Urophora populations below a threshold necessary to effectively control spotted knapweed. Received: 04 May 1999 / Accepted: 14 August 1999