Betty the Boozing Bee: Toward a Social Insect Model of Alcoholism

Methods

Summary

We will confined individual bees to a fully automated shuttle box (an apparatus where an individual honey bee can move between two compartments.) The basic experimental design will consist of Groups of bees (20 subjects each) placed individually in the shuttle box where one side contains bee food and the other contains either 0% ethanol (water only), 2.5%, 5% or 10% ethanol. The ethanol solutions are diluted with water. The bees will be confined for 10 days with observations on survival made each day. The data recorded include number of crossings and time spent on each side of the shuttle box. The shuttle boxes will be placed in a state of the art incubator to investigate the effect of temperature, light, and humidity on ethanol induced behavior. It is this incubator that I am are requesting funds for. After 10 days, the circadian rhythm and mortality data will be gathered. For two additional days (days 11 and 12) the ethanol will be replaced with water and observations made on tolerance and dependence.

Based on the results of our initial experience we will repeat the experiments for replication and extension into the area of learning and memory. After 10 days of ethanol exposure, we will conduct learning assays based on the proboscis extension response (bees will readily extend its proboscis to an odor signaling a brief sucrose feeding), the ability to avoid aversive stimulation, and in a situation where bees fly to a target where they are trained to push a cap to uncover a hidden food well.

Challenges

We are faced with two challenges. The first is establishing appropriate environmental conditions to keep the bees alive in an indoor environment. The incubator we have requested funds for will solve this problem. The second challenge is that the bees may not be able to chronically consume ethanol. In regards to the latter challenge our past field experiments suggests that bees will self-administer ethanol treated sucrose over several days.  In regards to the research design, we do not expect any additional significant challenges as we have published over 10 peer-reviewed papers on this topic and are well familiar with the methods and techniques – several of which we have developed in our laboratory.  

Pre Analysis Plan

For a number of years we have used a new method of analyzing data known as Observation Orientated Modelling (OOM) (Grice, 2011). OOM is a departure from traditional null hypothesis significance testing (NHST) and has several distinct advantages. These advantages include avoiding assumptions about normality, continuity, and homogeneity. In addition OOM facilitates abductive inferences and has the ability to test multiple hypotheses by comparing the observed data to competing theory-based predictions. As OOM does not rely on a normal distribution, alpha levels become meaningless as the emphasis is on explanation and replication. For a detailed comparison between OOM and NHST, see Dinges et al., (2013).