Evaluation of time-average dispersion models for estimating pheromone concentration in a deciduous forest

The Sutton and more recent Gaussian plume models of atmospheric dispersion were used to estimate downwind concentrations of pheromone in a deciduous forest. Wind measurements from two bivane anemometers were recorded every 12 sec and the pheromone was emitted from a point source 1.6 m above ground level at known rates. The wingfanning response of individually caged male gypsy moths (Lymantria dispar) at 15 sites situated 20 to 80 m downwind was used to monitor when pheromone levels were above threshold over a 15-min interval. Predicted concentrations from these Gaussian-type models at locations where wing fanning occurred were often several orders of magnitude below the known behavioral thresholds determined from wind tunnel tests. Probit analyses of dose-response relationships with these models showed no relationship between predicted dose and actual response. The disparity between the predictions of concentration from these models and the actual response patterns of the male gypsy moth in the field was not unexpected. These time-average models predict concentrations for a fixed position over 3-min or longer intervals, based upon the dispersion coefficients. Thus the models estimate pheromone concentrations for time intervals appreciably longer than required for behavioral response.