>Thirdly, the haplodiploidy hypothesis only works if all sisters share the same father and if a queen is biased to produce more daughters than sons. The sex ra

>Thirdly, the haplodiploidy hypothesis only works if all sisters share the same father and if a queen is biased to produce more daughters than sons. The sex ratio doesn't actually seem like a problem for the theory, because it sounds like for a worker bee, the relatedness of the marginal sibling is 5/8 in expectation, vs. 1/2 relatedness of the marginal offspring. I think you also have to discount the relatedness into the future. If the colony you are born into is already established, your 5/8 related marginal sibling has a much higher likelihood of survival than your 1/2 related marginal offspring when you take into account the risk of breaking from the colony and starting your own. That probably goes some way to explaining the first problem of multiple fathers. Marginal half-siblings are only 1/4 related to the worker, but they may have a greater chance of survival.