AbstractQuantification of interference with biological control agents can provide support for anecdotal claims of success or failure of agent establishment and efficacy. This study was initiated because of observed predation of cinnabar moth larvae by carpenter ants when releasing larvae for the control of tansy ragwort, an invasive plant in Montana. Biotic and abiotic factors were compared among three sites with historically variable moth population establishment. Two experiments were developed to (1) observe and document insect activity, predation, or disappearance on tansy ragwort stems either protected or accessible to ants; and (2) quantify the effects of ant exclusion on herbivory of tansy ragwort. Site comparisons indicated that ant colony density was highest at the driest of three sites, and, interestingly, no ant colonies were detected at the site with higher observed numbers of moth larvae and adults and lower densities of tansy ragwort. Available substrate (logs and stumps) for ant colonization did not differ between the three sites. In the ant exclusion experiments, a larger number of larvae were missing on plants accessible to ants (63%) compared with plants where ants were excluded (39%) after 36 h. Direct observation of predation of larvae by carpenter ants accounted for 9% of missing larvae on stems accessible to ants. Larvae were able to consume 81% of original flowers or buds on ant-excluded stems, compared with 18% consumption on ant-accessible stems, suggesting that ant predation could limit the efficacy of cinnabar moth larvae. These results provide one of many possible explanations for the anecdotal observations of large, persistent populations of cinnabar moths in moist areas. This work emphasizes the importance of post-release observation and monitoring to detect and, ideally, quantify factors to support anecdotal perceptions regarding the fate and subsequent efficacy of insect biological-control agents.
