Insects and influence: Interaction between behavior and physiology

Models representing processes that affect physiological development, phenology, and vectored-disease epidemiology are used to study and predict insect systems. Physical and physiological processes are often chosen for modeling because they respond to abiotic inputs with relative consistency, and enable applications such as weather-based forecasting of insect dynamics. Models of these processes most often relate causes to effects deterministically using a function. Uncertainty is then projected around estimates as the result of error in model parameter estimates (error in the model). In practice it is also assumed that the difference between recorded values of environmental variables and values experienced by modeled organisms or processes is zero, or is random (error in the data). Ironically, parameter estimates for physiological models such as those for arthropod vector development are often accurate and precise, whereas there is appreciable error in the data used to drive these models when implemented. In this presentation I will focus on modeling processes affected by adaptive response to environmental variation, demonstrating how models can be formulated to implicitly reflect behaviors. I will extend the application of this modeling approach to show how it can improve prediction of organisms' response to climate change. Examples of models and approaches emphasizing prediction will be discussed using necrophagous flies as a study system. A theme common in each modeling effort is the incorporation of factors that themselves determine environmental influence on processes, and the general approach is to combine physiological and behavioral/regulating functions through convolution to predict outcomes.