RELATIONSHIP BETWEEN CUCUMBER BEETLE (COLEOPTERA, CHRYSOMELIDAE) DENSITY AND INCIDENCE OF BACTERIAL WILT OF CUCURBITS

Field experiments were performed in 1993 and 1994 to establish varying densities of striped Acalymma vittatum (F.) and spotted, Diabrotica u ndecimpunctata howardi Barber, cucumber beetles in spring-planted cucu mber plots, and to determine the incidence of bacterial wilt symptoms in each treatment. These data were used to develop regression models t o examine the relationship between beetle density and the severity of bacterial wilt disease. Cucumber beetle numbers peaked during the firs t 2 wk after colonization of plants, with a rapid decline in numbers t hereafter Bacterial wilt symptoms were first observed 12-14 d after pe ak numbers of beetles were recorded. Critical point linear regression analyses of both year's data demonstrated a strongly Linear and positi ve relationship between cucumber beetle density and tile severity of w ilt symptoms. No significant improvement was detected by fitting quadr atic models. Beetle numbers recorded two weeks after colonization of p lants provided the highest r(2) values when regressed on die percentag e of wilted vines (final disease rating). Covariance analysis determin ed that regression slope values were not significantly different betwe en years; therefore data for both years were pooled and a generalized model was developed. Substituting area under the disease progress curv e values for the percentage of wilted vines per plant values resulted in a higher coefficient of determination (P = 0.0001, r(2) = 0.73, slo pe = 70.3, intercept = 92.6), indicating that use of the progress curv e, which accounts for disease development over time, provides a better recession fit than does use of a variable that reflects measurement o f disease incidence at discrete intervals. The regression analyses als o indicated that even low numbers (1 per plant) of beetles during the Ist few weeks after colonization of plants will result in wilt symptom s.