Evaluation of sequential sampling plans for the larger grain borer (Coleoptera: Bostrichidae) and the maize weevil (Coleoptera: Curculionidae) and of visual grain assessment in West Africa

Repeated sampling of rural maize stores in Benin was conducted to evaluate published parameters of a sequential sampling plan for a negative binomial distribution to determine pest status for Prostephanus truncatus (Horn) and Sitophilus zeamais Motschulsky. A computer program was used to rerandomize the data and evaluate the effects, in terms of average sample number and error rates, of different sampling plan parameter values. With respect to P. truncatus, lower and upper thresholds of 0.2 and 1.0 insects per ear and parameter values of k = 0.2 and alpha = beta = 0.1 were found to be adequate. With respect to S. zeamais, lower and upper thresholds of 10 and 20 insects per ear and parameter values of k = 1.0 and i alpha = beta = 0.1 were found to be adequate. Simplified sampling rules were proposed in which 11 ears should be sampled and if no P. truncatus are found, the population is low; otherwise the Wald plan should be followed. Owing to the lower per capita rate of damage, effective simplified sampling rules for S. zeamais were difficult to construct. An evaluation of the visual assessment scale rising whole ears showed that a visual scale estimating percentage damage rather than percentage loss, might be easy to construct and preferable for traders. Further work is needed to improve the usefulness of the visual scale in pest management decision support.