Drosophila are poikilotherms, which means they rely on the external environment in order to control their body temperature (i.e. they are ‘cold-blooded’). This means they are sensitive to the ambient temperature of their environment, and use behavioral strategies to optimize their body temperature. It’s important for them to be able to regulate their temperature since it promotes survival (extreme temperatures cause death) and affects metabolism, growth, movement, responsiveness and reproduction.
The effects of temperature on Drosophila have been studied basically since Drosophila has been used as a model organism. Harold Plough published that temperature affects the rate of chromosomes crossing over in Drosophila in 1917. I still think it’s amazing what they could do with careful observation of chromosome squashes. Temperature is considered a major stress that can drive evolutionary adaptation in flies (Parsons, 1960), and we know it has strong effects on where flies lay their eggs (Fogleman 1979) and other mechanisms that affect reproduction like diapause, which is a form of reproductive arrest at lower temps (Kimura 1988).
Thermosensation was one of the first behaviors that Seymour Benzer, one of (if not ‘the’) founder of Drosophila neurogenetics, studied. In “Behavioral genetics of thermosensation and hygrosensation in Drosophila” (1996) Sayeed and Benzer created a simple temperature (and humidity but more on this later) choice test, then screened 55 mutants to identify the bizarre mutant which doesn’t avoid hot temperatures. According to FlyBase bizarre still isn’t mapped. If you want to learn some of the mechanisms of thermosensation, a recent open access review “Thermosensation and Temperature Preference: From Molecules to Neuronal Circuits in Drosophila” by Chia-Lin Wu nicely covers the circuits and genes implicated in thermosensation in Drosophila.
So we know that flies are sensitive to ambient temperature, but does this affect whatever behavior we are measuring (courtship, locomotion, memory, etc)? In my lab we have observed that temperature has the most drastic effects on general activity levels and olfactory preference. There’s a goldilocks phenomenon where too cold or too warm slows them down and can create variability. Temperature can be used as reinforcer to drive behavior like in heatbox learning (Putz & Heisenberg, 2002), and to understand the mechanisms of place memory (Sitaraman et al. 2008; Ofstadt et al 2011) suggesting they definitely have a ‘happy’ or preferred range. Melissa and Troy Zars (2006) found that temperatures above 24°C have negative reinforcing properties in the heat-box paradigm. Which is interesting because most of us fly people keep our incubators at 25°C (FYI - we keep ours at 24°C).
My recommendation for performing experiments at ‘room-temp’ is to try to keep the range between 22°C-24°C. We find the most consistent behavior in that range. That being said, for very robust behaviors like the type you would perform in a laboratory class, we’ve found that we get pretty good effects when the room temp ranges from 18°C-26°C.
Sometimes you need to do experiments in higher temperatures for the thermogenetic tools you are using (for example, 30°C for shibirets or TrpA1 or Gal80ts experiments). It’s important to never consider the Gal4<UAS whose behavior was run in a 24°C environment as an appropriate control to Gal4<UAS whose behavior was run in a 30°C environment. This is because there’s a good chance the behavior you are measuring will be affected by temperature independent of the effect of blocking neurons or reducing gene expression. The most rigorous thing to do is run your UAS/+ control, Gal4/+ control and Gal4>UAS experimental line at both temperatures (usually 22°C and 30°C).
Since temperature seems to be so important, what is the best way to control temperature in your lab environment? Well, you can perform your experiments in a room that has temperature tightly regulated (which is often a room incubator that can be rather noisy, smelly, stuffy, and super expensive to build and maintain). We prefer a simpler solution. We build behavior boxes out of corrugated plastic sheets (like cardboard but plastic) and extruded aluminum railing (80/20). We mount heatable ceramic tiles (used most commonly for chick hatching) to the walls or ceiling of our behavior boxes, and use computer fans to help circulate the warm air. You can see some pics of different boxes on our Virtual Tour.
Do you have questions, helpful advice or anecdotes? Please post in the comments below!