Hacker News Reader

I wonder if they've done any experiments yet as to whether the birds actually can detect that. It definitely seems possible to do such an experiment.

I do think it's odd that in the article they just say the birds "must" be able to detect it, without any evidence that they actually can.

View on HN

For example you could have three boxes, only one of which has food in it. Paint that box with UV fluorescent material like that which is on their beaks and repeatedly show the bird where the food is and let it eat it. Then stop showing it and let it go to the box on its own. Then start switching around the location of the boxes and show the bird which box has food in it. Then keep switching the location of the boxes but don’t show the bird where the food is. Do this in all of the different lighting conditions that the bird would typically come across in nature. See if under any of these conditions the bird can consistently identify which of the boxes has food in it. If it can and if it’s not due to smell or sound then it has to be because of what it can see. Ensure that other than the fluorescence there is no other visual distinction that makes the box identifiable. Do this by making new boxes every now and then and replacing the old boxes with those.

In order to eliminate the possibility of sound or smell being a factor we could make three new boxes, one of which is fluorescent and the other two not as before but now instead of putting the food in the box we keep it in a separate bag and we reward the bird for going up to the fluorescent box by feeding it from our bag.

Bluetits (parus caeruleus) are sexually dimorphic under UV light only, and apparently someone did some experiments on this too:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1688915/

Reply on news.ycombinator.com