Many invasive ants are characterized as both unicolonial, and disturbed habitat specialists. However, so far unicoloniality of invasive ants has received far more attention than their dominance in disturbed habitats. Since Tsuji and Tsuji (1996, Oikos) I have argued that this phenomenon can be explained by their potentially high r (intrinsic rate of natural increase), a life history trait that is particularly advantageous in non-saturated environments. An alternative explanation to the success of invasive ants in disturbed habitats is the removal of native ants by humans. The latter hypothesis makes the implicit assumption that ant communities are at a non-stable equilibrium in a saturated environment. I propose a method that can distinguish the two hypotheses, and apply it to ant community data in Yanbaru forest in Okinawa. Data support the prediction of the high r hypothesis, namely that native ants will begin to dominate again when disturbance ceases.