Regulatory agencies assess risks to honey bees from pesticides through a tiered process that includes predictive modeling with empirical toxicity and chemical data of pesticides as a line of evidence. We evaluate the Varroapop colony model, proposed by the US EPA for pesticide registration evaluations, by simulating hive cohort dynamics with Monte Carlo simulations and sensitivity analysis techniques. We use a modified version of Varroapop, Varroapop+Pesticide, to predict population growth and behavior temporally in three pesticide exposure scenarios (i.e. foliar application, seed treatment and soil application) and a baseline scenario. We use linear approaches to assess first-order parameter sensitivities which allows us to determine how variance in the output is attributed to each of the input variables across different exposure scenarios; in addition, the daily resolution of the model allows us to conditionally identify sensitivity metrics. Our results highlight important contributors to honey bee exposure and risk, identify in-hive parameters that may require additional research, inform the calibration of sensitive parameters, avoid overparameterization, and assess the relative importance of model submethods of hive population dynamics.
Results/Conclusions
By examining partial correlation coefficients from day to day, we were able to identify conditional model variability and attribute these sensitivities to seasonal and life history dynamics. Further investigation with daily conditional sensitivity analysis reveals different pathways of pesticide exposure within the VarroaPop+Pesticide model. Foliar spray application rate and adult contact LD50 become critical parameters for colony size after exposure which indicates mortality of foraging individuals via direct contact. Nectar load, nectar consumption, and nectar foraging trips become critical parameters for colony survival in the seed and soil scenarios which indicates mortality of honey bees via indirect oral exposure to pesticides. We also observe a minimum colony survival threshold in our Monte Carlo simulations. Simulations with the model indicate queen strength and forager lifespan are consistent, critical inputs for colony dynamics in both the non-exposed and exposed conditions. The combined influence of queen strength and forager lifespan creates a minimum boundary condition for model parameterization and colony survival which fluctuates according to other influential parameters that arise with each pesticide application method. Sensitivity analysis as conditioned by day also reveals that the relative importance of queen strength and forager lifespan fluctuates throughout the simulation period and yields to toxicity and foraging parameters following pesticide exposure.