How you design the behavior arena (ie the little chamber that you place the flies to study fly behavior, will affect the behavior that you see. For example, when you reduce a 3.81cm diameter circular arena to 1.27 cm diameter, male but not female flies increase their locomotion (Xiao and Robertson, 2015). Overall, it’s very important to think about arena size when you design your experiment. Below are some tips for designing your behavior chambers. How to obtain the materials and build the arenas will be in another post. Sexual dimorphism and behavior will also be in another post.
How high should my behavior arena be? If you are tracking flies and don’t want your flies to be able to walk on the ceiling and over each other, use a 2.3-2.5 mm height. To prevent flies from climbing on the sides and ceiling of the chamber, use teflon paint (more on this later).
What area of behavior arena should I use? The area of your optimal behavior chamber will depend on the type of behavior you want to perform. For example, courtship measures (like number of wing extensions, time spent chasing, time spent licking, etc) are very different in a traditional small courtship chamber compared to a larger arena (check out three CSHL protocols from von Philipsborn, Shohat-Ophir and Rezaval for more details: Measurement of Drosophila Reproductive Behaviors, Single Pair Courtship and Competition Assays in Drosophila, and Courtship Conditioning/Suppression Assays in Drosophila). The Drosophila FlyBowl is 127 mm in diameter (Simon & Dickinson 2010), and we the FlyBowl in our lab so we can take advantage of the BABAM (Browsable Atlas of Behavior-Anatomy Maps) Database (Robie et al, 2017).
What density of flies should I place in an arena? The behavior you will see will also depend on the number of flies in the arena (Rooke et al, 2020). For the group behaviors we measure in the lab, we prefer about 10 flies in a ~40mm diameter chamber. We find we can put up to 30 flies in that size of an arena, but it gets difficult to track individual flies (ie soooo many tracking errors) and the activity increases because of the amount of social interactions that occur. If you want to use densities of around 30 flies we recommend going up to an arena about 90mm in diameter (FlyBowls work well with this density).
What shape of behavior arena should I use? Again, the shape of the arena that you use will depend on what type of behavior you want to see. Over the years, we have moved towards chambers with rounded edges (circles as opposed to squares or oblong ellipses as opposed to rectangles) because if you are tracking flies for any length of time, the flies like to hang out in corners. By rounding out the corners, you can increase the activity of the flies and reduce the number of tracking errors you get by flies bunching up in small spaces. For single fly assays, we don’t like using runways longer than 100mm. Some arenas (like the FlyBowl) have gently sloping floors which restricts the space near the edges and encourages flies to spend more time in the center of the arena. This is helpful because flies like to spend a lot of their time near edges.
In our experience, simple is best. When we’ve designed complex arenas with different ‘rooms’ for example, the data is always variable and the effect can be difficult to reproduce. My recommendation is to start with something simple with rounded edges and no moving parts.
What arena dimensions have you found work best for you?