If the least frequent sub families are more likely to be chosen as potential queens I wonder what happens with the typical grafting process where the grafts are offered to a cell raising colony which is unrelated to the queen the grafts were taken from.