Binomial sequential sampling plans for late instars of European corn borer (Lepidoptera: Crambidae), corn earworm...

Late-season infestations of European corn borer, Ostrinia nubilalis (Hübner), and corn earworm, Helicoverpa zea (Boddie), were sampled to develop binomial sequential sampling plans for larval infestations and damaged kernels in sweet corn, Zea mays L., ears, near harvest. Fields were sampled to obtain a range of larval densities likely to be encountered over a range of infestation levels and field conditions. Binomial sampling plans were developed for O. nubilalis larvae, H. zea larvae, O. nubilalis, and H. zea larvae combined, and for damaged sweet corn kernels. Observed densities ranged from 0.01 to 4.40 larvae per ear for O. nubilalis, 0.005-1.62 larvae per ear for H. zea, and 0.004-36.12 damaged kernels per ear. Results of resampling analyses, based on the proportion of ears infested with one or more larvae, or damaged kernels, indicated an average sample size of 34-37 ears was necessary to classify whether larval infestations, or the incidence of damaged kernels, exceeded 5%. Two operating characteristic curves are presented for each of the four sampling plans. Initial results, with upper bounds of 0.10, and alpha (type I) and beta (type II) error rates at 0.10 and 0.05, respectively, resulted in a 90% probability of making the correct management decision at infestation levels >10%. To improve performance of the sampling plans, we modified the binomial plans by reducing the upper bound to 0.075, while maintaining the same error rates. This plan resulted in a higher probability (>95%) of making the correct management decision to reject a sweet corn load when infestation levels are >10%.