Variable-Intensity Sampling: A New Technique for Decision Making in Cabbage Pest Managementl

J. Ecan. Entomal. 76: ]39-]43 (]983) A variable-intensity sampling method was developed to efficiently monitor all cabbage pests in commercial fields in New York while concentrating on the most important pests, Pieris rapae (L.) and Trichoplusia ni (Hiibner). The field is walked in a V-shaped pattern which is divided into 10 segments. Along each segment, one to four randomly selected cabbage plants are inspected, the number depending on an estimate of population density of P. rapae and T. ni derived from previous segments. Computer simulations of this procedure were run, using random numbers from the appropriate negative binomial distribution, and the 95% confidence intervals calculated contained the true mean with frequencies between 93 and 97%. In a commercia] cabbage field, a comparison of estimates of the means obtained by variable intensity sampling and fixed sample sizes of 40 and 264 plants revealed only small differences. portant cabbage pests, T. ni and P. rapae (Andaloro et al. 198]). A sampling procedure was developed and tested based on the aforementioned objectives and the results are reported here. Materials and Methods Sampling occurred along a V-shaped transect, extending from one comer of the field to the middle of the opposite side and back to the other comer. One to four randomly selected cabbage plants were inspected along each of 10 equal segments of the transect. As sampling proceeded, the number of plants sampled along the next segment varied, depending on estimates of T. ni and P. rapae population density derived from previous segments. Insect pests, other than T. ni. and P. rapae, were noted on the plants examined. We conducted an extensive visual survey for diseases and weeds while walking along the transect. Pest management should rely on some method of estimating pest population densities for treatment decisions. Sampling procedures that utilize a fixed sample size are inefficient when the pest population density is either very low or very high, because the sampling intensity is greater than necessary to make an estimate for a control decision. Sequential sampling, in which sampling is terminated when the pest population density is classified as either above or below a critical density, is more efficient (Onsager 1976). Sequential sampling plans for cabbage looper, TrichopLusia ni (Hubner), (Shepard 1973) and imported cabbageworm, Pieris rapae (L.), (Harcourt 1966a), have been developed for treatment decisions on fresh market cabbage. However, treatment decisions based on sequential sampling of insect pests alone can be unrealistic. If the insect pest population is within a critical range, the treatment decision is difficult, because additional variables such as present and expected growing conditions, previous crop stress, crop growth stage, expected crop value, and environmental factors that affect the pest population also must be considered. The treatment decision can be made easily on the basis of insect pest population alone if the density is either very low or very high. A sampling procedure should result in the most precise estimate when many variables must be considered in the treatment decision, but for efficiency, the estimate can be less precise when the treatment decision can be made solely on the basis of insect pest population density. An important objective in cabbage pest management is the early detection of other less common but potentially serious insect, disease, and weed pests. This requires an extensive survey of the entire field, which cannot be done if sampling is terminated quickly based on a sequential sampling decision. A need exists for an efficient sampling plan for any pest that might require control measures, but focuses on information for treatment decisions for the most im40