Working with Aaron Tarone of Texas A&M Entomology and members of his group, we have been developing insect phenology models that account for thermoregulatory behavior. These models are intended for use in predicting vector and pest dynamics, and in retrodicting events of forensic importance that can be indicated by insect development.

We’re happy to have some of this work recently published in Frontiers in Ecology and Evolution: https://static.frontiersin.org/articles/10.3389/fevo.2022.837732/full

The rationale of this work has been that organisms capable of behavior will behave in ways that are on average beneficial. Insects have been shown to “have” what we call “preferred temperatures.” The mechanisms related to this are complex, but they don’t need to be clarified before we can improve predictions by implicitly incorporating behavior into models. Our improvement transforms distributions of environmental temperature into distributions of insect temperatures, through a function that involves environmental temperature and its variance, and insect preferred temperature. It is a simple convolution, wherein development rate is a function of insect body temperature, which is a function of environmental temperatures and insect behavior.

Here’s how it looks:

The graph shows temperatures through time: ambient air (which is what many phenology models use as input), colonized flesh (because this work was done on necrophagous fly larvae), and predicted insect temperature using our model. I’ve arbitrarily chosen 27 °C as a preferred temperature. Blue bands around the ambient temperature trace are to suggest temperature variation across space.  Predicted larval temperatures lie between ambient and preferred, determined jointly by the difference between ambient and preferred and the variance around ambient.