A computational approach to tracking Western Honeybee tasks based on age


Apis mellifera,
or the Western Honeybee, is the most common honeybee. Past research has indicated the Western Honeybee takes on different responsibilities within the hive as they age (referred to as age-based polytheism). The system does depend on context, meaning if older bees die out, younger bees will take on those responsibilities faster or vice versa. The authors of this article developed a computer model to estimate when members of a brood cohort (i.e. a group of bees whose eggs were laid around the same time and hatched together) will most likely begin undertaking specific tasks.

The model has multiple advantages. Researchers can track how exchanging brood cohorts between colonies or brood acute exposure to pesticide affect the rate at which the brood undertakes specific tasks. The model is also much less costly than large field experiments and avoids statistical “fishing” pitfalls. The user inputs data for the model to work including, how much of the brood comb is filled by eggs, larva (capped and uncapped). The authors mentioned in the paper that future development could allow this step to be completed by image recognition.

The model accounts for error and uncertainty by running multiple simulations using developmental rates gathered from samples provided in a given location (the authors used Alabama and Georgia in this model). The model produces a distribution of ages for each cohort. The model can be used to predict “Cohort Contribution”, or how much of the cohort will undertake a given task.

The authors produced this model for the purpose of guiding experimental work, understanding how stressors impact brood development, and to measure the effects on individual performance following those stressors. Upcoming work from researchers in at University of Georgia, Auburn University, and University of Delaware used this new toolkit for understanding how pesticides affect the abilities of bees to perform their tasks, and how beekeeper interventions can compensate for that.

To read the full article please click here. 

Corresponding author: Lewis J. Bartlett, lewis.bartlett@uga.edu

Summary by Jannah Zinker