Designing a Feeder

Right now we’re working on designing an artificial feeder that will be able to keep track of when bees visit it, and this week we brainstormed features it needs to have. We started with a very general idea of what we want the feeder to look like: It needs to have some kind of central container that holds an artificial nectar solution, and a set of feeding stations around it so that bees can access the nectar. Each feeding station will have some kind of artificial flower to attract the bees. With this premise in mind, we made a list of questions that a feeder like that could help us investigate. Essentially, we’re interested in the arrival patterns of bees to the feeder, and how those patterns are affected by different circumstances. An arrival pattern is determined by questions like:

• When did the first bee arrive at the feeder?
• Do bees arrive at an increasing rate throughout the day?
• Do the bees arrive in clusters, or are they spread out?
• What time of day do most bees visit?
• How long does each bee stay?

We want to find out how the arrival patterns are affected by things like:

• Changing the nutrient mixture of the nectar
• Changing the scent of the nectar
• Changing the appearance of the feeder
• Adding features that the bees perceive as dangerous or unpleasant

We also want to investigate whether the bees show any preference for certain stations on the feeder (stations that have or have not been visited by other bees, that have different appearances, etc). We hope that by investigating arrival patterns, we can learn something about how the bees communicate with each other about food sources.

With this list of questions in mind, we started to define physical and data-collecting features that the feeder needs to have. As far as data-collecting goes, the feeder needs to notice when a bee arrives and what station it visits, and record what time it got there and how long it stayed. The feeder needs to be able to store data on site (without being connected to a computer), in a form that can be uploaded to a computer later.

Our current plan for the electronics is to use an Arduino and a data logging shield. Each feeding station will have a photointerrupter gate on it, which will signal the Arduino whenever a bee lands or takes off. The data logging shield will time-stamp the information and record it on an SD card as a list of when bees visited each station and how long they stayed, so that at the end of the day we will be able to upload that data to a computer and view it.

To collect meaningful data, the feeder needs to be attractive to bees. The artificial nectar will be made of sugar water infused with some combination of peppermint, aniseed, and clove bud oil, because bees are attracted to both scents and nutrients. Each feeding station will have an artificial flower around it, to attract the bees visually. The flowers must be detachable, so that we can investigate whether different appearances influence arrival patterns. Each station also needs to be clearly labelled, so that we can match up the data to the stations. Refilling a feeder can be disruptive and temporarily scare away the bees, so we will get more accurate data if the feeder can hold enough nectar to last for the whole day. We also want to prevent other animals (like ants and hummingbirds) from drinking the nectar. Since we might investigate the effects of different nectar compositions, we don’t want residues from one mixture to stay on the feeder when a different mixture is being tested. The scented oils are very concentrated, so a slight residue could have a strong enough scent to attract bees. Therefore, the feeder needs to be easy to clean and have interchangeable parts (so that different feeding trays could be used for different mixtures).

Now that we’ve got some idea of what we want, it’s time to start diving into the actual design. And of course, we’ll keep you up to date on what we come up with!