Juergen Heinze has a must-read piece in the latest edition of Myrmecological News about how ant colonies are not often simple families as we like to think of them:
Abstract: The social systems of ants are far more variable than has traditionally been believed. In addition to variation in queen number and queen mating frequency, recent research has documented such bizarre phenomena as the parthenogenetic production of females from unfertilized eggs or genetic caste determination. All these affect the genetic structure of ant societies, and it appears that in a large percentage of species colonies do not consist of a single, singly mated mother and her sterile worker offspring. Though it has long been known that kin selection for reproductive altruism can work well without a relatedness value between workers of 0.75, the recent upsurge of discussions about the role of relatedness in kin selection theory may have confused both myrmecologists and non-specialists. The aim of this review is to give an overview of the large range of ant reproductive systems and to correct some misconceptions about the role of the magic value 0.75 in kin selection theory.