Patterns in Resource Partitioning by Insect Biological Control Agents of Yellow Starthistle (Centaurea solstitialis L.) in Washington

摘要

Entire yellow starthistle, Centaurea solstitialis L. (Asteraccae), plants were harvested from three southeastern Washington sites in September 2004 for seed head dissections to assess intra-plant patterns in the spatial distribution of the introduced false peacock fly, Chaetorellia succinea (Costa), and peacock fly, Chaetorellia australis Hering (Diptera: Tephritidae). Thirty plants were partitioned into three height classes and collected at each site, spanning the longitudinal range of yellow starthistle in Washington. Seed head canopies were stratified into three vertical sections and individual heads were dissected to determine bioagent occupation. Logistic regression analysis using a binary logit model was performed on data collected from each of the three sites separately due to varying site effects. Explanatory variables included plant height, plant width, plant canopy stratum (apical, mid, basal), and presence of other bioagents within a head. 'Other bioagents' detected include the yellow starthistle hairy weevil, Eustenopus villosus (Boheman) (Coleoptera: Curculionidae), the yellow starthistle bud weevil. Bangasternus orientalis Capiomont (Coleoptera: Curculionidae), the yellow starthistle flower weevil, Larinus curtus Hochhut (Coleoptera: Curculionidae), and the yellow starthistle gall fly, Urophora sirunaseva (Hering) (Diptera: Tephritidae). Using Chaetorellia spp. presence its the response variable, models were significant for the western-most site in Klickitat Co. (chi(2) = 35.66, df = 5. P<0.0001, n = 2,783), and the eastern-most site in Asotin Co. (chi(2) = 17.93, df = 5. P = 0.0030, n = 939). but not for the central site in Benton Co. (chi(2) = 9.44. df = 5, P = 0.0929. it = 1 MO). The one consistently significant site-wide predictor of Chaetorellia spp. occupation was the presence of 'other bioagents.' The frequent co-occurrence of Chaetorellia spp. and, in particular, E. villosus at the sampling sites suggests potentially competitive interactions between these insects.