Sensitivity Analysis of Deterministic and Stochastic Simulation Models of Populations of the Sheep Blowfly, Lucilia sericata

Abstract

Using empirically derived relationships between temperature and development rate, deterministic and stochastic simulation models were constructed to predict the seasonal pattern of abundance of the sheep blowfly, Lucilia sericata. The number of day degrees accumulated each day by a cohort was calculated from the daily temperature pattern and the base temperature threshold for that stage. The diurnal temperature pattern was described by a simple sine curve relationship, based on the maximum and minimum temperatures. The stochastic model uses a Monte-Carlo simulation technique to assign random development rates to each life cycle stage generated from a Weibull distribution fitted to observed variation in development rate for each stage. The deterministic model typically predicts four discrete waves of emergence of L. sericataadults each season, with the fifth generation limited by the inset of dispause. The stochastic model resulted in a similar predicted population but with less discrete generations than the deterministic model and it more closely resembled the pattern of blowfly abundance observed in the field. Sensitivity analysis showed that both a reduction in the base temperatures and the number of day degrees required for life cycle development increased the daily rate of development at a given temperature and resulted in higher mean numbers of flies present each generation. However, the pattern of blowfly abundance observed was more sensitive to variations in base temperatures than day degree requirements.