64% of the time 10 bees chose 0 is statistically significant but not with a low p-value.

You’d need to invest a lot more time and effort into proving this to get the P-value lower and I don’t expect many institutions are willing to train 1000 bees to try and figure out if they get 0 or not.