Data scientist at Port Jackson Partners in Sydney, Australia. My PhD was in computational biology. In my spare time I write about medical research at BioSky.co.CVAbout
Sometimes missing data can tell you its own story, while other times missing data can lead you in completely the wrong direction. As I’ve been working on a PhD studying honeybee behaviour, I thought I’d share a couple of articles about bees and how important missing data can be.
This first article is the story of how I was tearing my hair out in frustration trying to get tags to stop coming off my bees when something else entirely was going on.
As part of my research, I developed a system that enabled me to tag and track bees inside the beehive over an extended period of time. When I first started my tagging experiments, I was struggling to find a way to attach tags to the bees that would stay on for a couple of weeks. Eventually, I found a glue that I thought would work very well, but I kept coming back the next day and finding all my tagged bees were missing from the beehive.
Originally, I thought that the glue or the tags just weren’t good enough, but I always found it strange how the tags on 200 bees could come off so easily within a few hours. Eventually, I stuck a bucket at the hive entrance, and when I returned a few hours later, I found all 200 of my tagged bees were sitting in the bucket outside the hive.
I then put the bees back into the hive and watched them in the dark (our observation hive has a glass side). What happened next revealed the exact cause of all my frustration.
You see, the bees already in the beehive didn’t like the smell of the glue on the tagged bees, and were going up to my tagged bees, grabbing them by the wings, dragging them to the entrance, and flinging them out. The tagged bees would then normally die during the night and ants would drag them away, which is why I thought the tags were just coming off!
In retrospect, this behaviour made a lot of sense purely as a hygienic behavioural response – from the perspective of the bees already in the hive, the addition of unfamiliar bees that were in close proximity to the brood of the colony and that smelled funny must have been cause for considerable alarm!
Here was some footage I took of them dragging around some of my tagged bees within the hive:
So originally, the missing data (disappearance of tagged bees within the hive by the workers, and without the hive by the ants), led me to believe that yet again my tags and glue were simply coming off. It was a great lesson for me in how sometimes you need to double check your assumptions and have the patience to simply sit down and watch.
Finally, for those who might be curious about how I solved this particular problem once I understood it, I discovered that if I put my tagged bees on a frame in the incubator for several hours after I tagged them (until the smell was gone), they would be accepted into the hive.