Sweepnet Sampling for the Cereal Leaf Beetle, Oulema melanopus12

Two interrelated models were developed to predict cereal leaf beetle density from sweepnet catch. The mathematical model requires knowledge of sweepnet diameter, length of sweep, the proportion of the insect population that is in the net's path, and the probability of capture given that the insect is in the net's path. Difficulty in submodeling the last 2 variables led to the development of an alternative regression model. This model is specific to adult cereal leaf beetles and assumes a 15-inch net and a 5-foot sweep. It requires knowledge of wind speed, temperature, solar radiation intensity, and crop height. The regression has an r 2 of 0.87. Variance analysis showed that between-field variance is much greater than within-field variance and that the difference increases as density increases. Optimal allocation of sampling resources requires one sample per field and as many fields as possible with those resources. Normally if 60% of the fields of one crop in an area are sampled, then for that crop the standard error of the catch is within 10% of the mean. For detection purposes it is best to sample during peak larval densities to insure the greatest chance of a positive find. The sweepnet is an extremely efficient detection device down to densities as low as one insect per acre. Even if no cereal leaf beetles are caught in a region, probabilistic statements can be made regarding maximum likely densities.