Implications of individual variation in insect behavior for host specificity testing in weed biocontrol

Abstract

This study shows that individual behavioral variation is an under-recognised source of error that may affect the outcome of host range tests in a stenophagous species. Original specificity testing of the broom seed beetle, Bruchidius villosus (F.) (Coleoptera: Chrysomelidae: Bruchinae), a biocontrol agent for Scotch broom, Cytisus scoparius (L.) Link (Fabaceae: Genisteae), failed to detect its ability to oviposit in the field on a congeneric non-target plant, the exotic Cytisus proliferus L.f. (Fabaceae: Genisteae). These tests were repeated using individual beetles from the original UK collection sites and from New Zealand, 15 generations post release. In the original tests, low replication of small batches of females masked high levels of individual variation in oviposition preference. Although most beetles showed strong preference for the target weed, there was some indication that New Zealand beetles showed higher preference for the non-target than UK beetles.